Skip to navigation – Site map

HomeAnglophonia24Converging accommodation in the B...

Converging accommodation in the Black Country Dialect

Lyndon Higgs


This article illustrates and analyses various forms of grammatical accommodation in a corpus of recorded informal conversations and interviews with 22 Black Country (BC) dialect speakers and 3 Standard English (SE) speakers. The Black Country is an area of central England to the west of Birmingham. The variety of English spoken here contains many non-standard grammatical features, especially in the realisation of auxiliaries and pronouns. The paper begins with an analysis of the deliberate converging accommodation techniques used by the SE speaking interviewer, in an attempt to reduce the effects of the observer’s paradox during some of the recordings. Analysis of the data shows that this accommodation towards BC dialect was partial and occasionally inauthentic, yet it did indeed play a significant role in producing natural, authentic speech samples. Trudgill (1992: 7) comments that accommodation normally takes place during face-to-face intervention. However, this corpus contains examples of much more complex types of accommodation in reported speech. The paper analyses an extract where person X reports to person Y a conversation X had with person Z, in which Z reports to X a discussion she had with person A. Accommodation towards and away from BC dialect forms serves to disambiguate the situations of utterance involved, as well as clarifying relationships between the speakers. Finally, an example of what initially appeared to be the much rarer phenomenon of diverging accommodation is studied: a BC dialect speaker uses a SE feature (the third person subject pronoun she) in a context where the BC dialect variant (her in subject function) would be expected. Upon closer analysis, however, a second, more satisfactory, explanation is posited: the choice of pronoun (her or she) depends primarily on the relationship the speaker has with the referent of the pronoun in question.

Top of page

Full text


1This article examines converging accommodation phenomena in a corpus of recorded informal conversations and interviews with 22 Black Country (BC) dialect speakers and 3 Standard English (SE) speakers. The Black Country is a central England region found to the west of Birmingham, including the towns of Wolverhampton, Dudley, Walsall and Stourbridge. Although its exact borders are frequently disputed by local residents, it is generally agreed that its northern boundaries run along those of Staffordshire, while its southern limit touches those of Worcestershire. Directly to the west are South Staffordshire and Shropshire, and to the east are Birmingham and Warwickshire.

  • 1 BATH, STRUT, TRAP and FOOT are part of the lexical sets introduced by Wells (2002) to illustrate cl (...)

2Phonologically, the variety of English spoken here is generally described as being more northern than southern (for example, the BC vowels in BATH and STRUT are those found in the R.P. realisations of TRAP and FOOT)1. Although outsiders often confuse it with the nearby Birmingham variety, local residents are proud of its specificity, and there is a network of local performance artists who promote the Black Country dialect as well as the region’s history and culture. On a national level, however, this dialect, along with its neighbouring Birmingham variety, systematically scores low in popularity polls, whether they be academic, sociolinguistic or popular press studies.

3Accommodation theory, initially developed by Giles (1973), describes two distinct types of accommodation within the sociolinguistic framework: converging and diverging. Converging accommodation is generally understood to be the phenomenon whereby speakers make a temporary or long-term adjustment to their language variety (phonologically, grammatically, or stylistically), in order to approximate their language to that of their interlocutor’s. The reasons for this converging accommodation are many: either they see their interlocutor’s language as socially desirable, or they wish to show good will towards them, or they wish to identify with them. Converging accommodation may occur towards, or away from socially prestigious varieties, depending on the context. The rarer phenomenon of diverging accommodation, on the other hand, is the distancing of one’s language from that of the interlocutor, in order to disassociate oneself from them, or to assert one’s own identity, or to indicate social distance or disapproval. Again, diverging accommodation can imply a movement towards, or away from, a socially prestigious variety. Accommodation theory also explains that converging accommodation can backfire, so to speak, in as much as the speaker who wishes to express solidarity towards their interlocutor can, if not careful, overdo their accommodation (so-called over-accommodation) and be perceived as being condescending, despite their good intentions. This is typically the case when, for example, a Standard English speaker wishes to express solidarity with a non-standard dialect speaker, and deliberately (or subconsciously) tries to accommodate their standard dialect towards the non-standard variety. This aspect will be developed in more detail in sections 2.1 and 2.2.

4This article aims to analyse accommodation phenomena which occur specifically within the grammatical domain of the BC dialect. In order to do so, it was imperative to use a fieldwork technique that would provide totally natural speech samples. Early fieldwork techniques such as Labov’s (1966) and Trudgill’s (1974) had involved random sampling, meaning that informants in the community were chosen at random for the study and thus inevitably the fieldworkers were strangers to them. While this technique allowed access to more formal styles of speech, various researchers have noted shortcomings in the elicitation of casual speech. At best, a relatively formal conversational style was obtained. Gal (1979) noted that early attempts at interviewing Austrian informants produced a frustratingly narrow range of styles, despite efforts to encourage involvement and informality. Sutcliffe (1982) makes similar comments on his interviews with British Black English speakers. However, since Trudgill and Labov were chiefly concerned with phonological realisations, and not grammatical ones, this fieldwork technique did not prevent them from obtaining satisfactory results. In the present study, such a technique would not be appropriate, as informants would be extremely unlikely to produce a sample that contained typical grammatical features of the dialect, since generally speaking, grammatical dialect forms are heavily stigmatised and are thus avoided in relatively formal conversational styles. Clearly, then, another fieldwork technique was needed, and this will be described in 1.1.

5This article is structured as follows. After the description of the framework of the study, an outline of the BC dialect’s salient grammatical features will be given. Then several types of accommodation will be examined in the corpus of data collected on the BC dialect. Firstly, the converging accommodation techniques employed by the interviewer will be analysed. Next, an example of converging accommodation within the context of complex reported speech will be studied. Finally, an occurrence of what, at first sight, appeared to be an example of the much rarer phenomenon of diverging accommodation will be examined, before a more satisfactory explanation for this occurrence – i.e. the fact that the BC dialect has a third person binary politeness distinction that is not present in SE – is put forward.

1. The Black Country dialect: the framework of the study and salient features of the dialect

1.1. The framework of the study

6This analysis is based on a larger study (Higgs 2004), whose objectives were to describe the salient grammatical features of the dialect spoken by some inhabitants of the Black Country. Unlike Heath’s (1980) description of the variety of English spoken in nearby Cannock, phonological features were not studied. The frequency and variation of these salient grammatical features were analysed, taking into account external factors such as age, social class and situational context. A qualitative analysis of code-switching between Standard English and dialect was also undertaken.

7The framework of the research involved recording samples of twenty-five informants’ speech. The main objectives underlying the informant selection criteria were to cover a wide age range (the informants were aged between 7 and 77) and, most importantly, to obtain samples of natural informal speech.

8As explained in the introduction, restricting the data base to formal interviews with randomly chosen informants would clearly have been inefficient, since the results obtained would have contained far fewer dialect grammatical forms than in authentic natural speech: most informants would certainly have chosen to use SE forms in an interview with a stranger. Thus another technique was needed. Ideally, I wanted to record a series of natural conversations between two or three informants, in order to capture samples which truly represented the dialect vernacular. Being a native from the area, I took advantage of my so-called second order network contacts (friends or ‘friends of friends’) to find as many informants as possible. Most recordings were therefore made in a totally natural environment, either with or without my presence.

9In other cases, however, I did resort to interviews, simply because I did not have access to second order network contacts of a particular age-range, particularly elderly people. For this, I conducted very informal interviews in an old people’s day centre. In these cases, I took care only to choose informants who seemed particularly at ease in this interview situation. Moreover, an attempt was made to let the informant guide the conversation as much as possible. If necessary, I tried to evoke subjects of conversation which would typically interest the informant, frequently asking about family or ‘the good old days’. Before beginning the interviews, I spent two days at the local day-centre, chatting and helping to serve tea, and I made a deliberate effort to make my local origins known to the informants.

10There was a clear gender bias in the study, as 20 out of 25 informants were women, the reason for this being simply that women were more readily available through network contacts. According to Macaulay’s (1991: 16) method of social classification (based on occupation, education and residence), 23 belonged to the lower class and 2 to the middle class. The samples varied in length; the longest contained 8200 words (46 minutes), the shortest 500 words (5 minutes), the average size being 2500 words. The complete data contained 49000 words, roughly the equivalent of 4.5 hours of recording. The samples also varied in topic, but all shared the common feature of being natural informal speech. All samples were transcribed orthographically in their entirety, the informants were given pseudonyms, and all grammatical dialect features were coded (phonological features were not studied). Additionally, all Standard English elements for which a dialect form exists were also coded, thus allowing their frequency to be calculated. Typically, the dialect forms were given a code in capital letters (e.g. ‘BE’ for a dialect form such as I ay (= I’m not) and SE forms were given the same code in small letters (e.g. ‘be’), thus enabling ‘clusters’ of dialect forms in the transcripts to be highlighted easily. All in all, around 50 dialect features were coded, including auxiliaries, lexical verb forms, prepositions, pronouns, plural noun forms and determiners. The most important features are described in the next section.

1.2. Salient features of the dialect

  • 2 See Higgs (2004) for further information on the grammatical features of the dialect.

11A detailed description of all grammatical features of the BC dialect is clearly outside of the scope of this article. However, below are some of the main features that characterise it.2 Within the domain of verb forms, the auxiliaries DO, HAVE and BE all possess dialect variants, which occur either as a constant, or as a variable (i.e. sometimes realised as the SE equivalent). Table 1 summarises the dialect variants of these three auxiliaries, in present and past forms, with an example to illustrate each category.

12Table 1. Dialect forms of the auxiliaries DO, BE and HAVE




Affirmative DO

Identical to SE

I do know

Identical to SE

I didn’t know

Negative DO

I/you/(s)he/we/they DOW

(roughly the same diphthong as in know)

I dow go

I/you/(s)he/we/they DAY

or (s)he DON’T

I day go

Affirmative BE

I/you/we/they AM

(S)he IS

We’m going.

I/you/(s)he/we/they WAS

They was going

Negative BE

I/you/(s)he/we/they AY/ AINT/ INT

They ay/aint/int going

I/you/(s)he/we/they WONT/ WOR/ WASN’T

They wont/wor/wasn’t going

Affirmative HAVE

Identical to SE

We’ve seen him

Identical to SE

We’d seen him

Negative HAVE

I/you/(s)he/we/they AY/AINT/INT*

(*INT only 3rd person sing. and pl.)

I ay done it

Identical to SE

I hadn’t done it

13Tags associated with the modal MAY (and its past form MIGHT) are systemically HAD (I might as well wash it, hadn’t I?). Similarly, the semi-modals USE TO and OUGHT TO are constructed with an auxiliary HAVE: I hadn’t used to do it, had I? / You hadn’t ought to do that. In present continuous forms, a-prefixing sometimes occurs, whereby the prefix a-is added to the verb form: they’m a-awaiting (in SE: They’re waiting). This phenomenon is, however, only found in elderly dialect speakers’ speech. As is the case in most regional dialects of English, many non-standard past forms of lexical verbs exist: he’s stole it / he’s forgot it / last week he come. Extension of third person ‘s’ to first person forms is extremely common in the case of historic present narratives (and I says, what do you want, and he says, nothing), but also exists in other contexts, often indicating annoyance or regret (in older speakers only): I wants more days / We has to pay our taxes. Finally, within the realm of pronouns, thee and thou are still found – albeit rarely – in older dialect speakers’ speech. Moreover, the 3rd person feminine subject pronoun (she in SE) is realised as her (he told me her was coming) in the speech of many dialect speakers, both young and old. According to Trudgill (1990: 89), this so-called pronoun exchange is a relatively rare phenomenon existing in some dialects spoken in an area of South West England that encompasses Wolverhampton, part of the Black Country, at its northern limit. Hernández y Siebold (2010: 121) also points out that subject her is overwhelmingly a Midlands and South-West phenomenon.

2. The interviewer’s accommodation techniques

2.1. A quantitative study

14Despite being a native to the area, my main concern was that in the informal interview situations conducted in an old people’s day centre, the fact that I was a stranger, and that the variety of English I spoke was to a large extent SE, might affect the authenticity of speech obtained, and thus any meaningful statistical comparison with the other samples, obtained in a peer-group situation, would be jeopardised. For this reason, in the informal interview samples (hereafter referred to as group B), I decided to make a deliberate effort to accommodate my speech towards that of my interviewees. In the other recordings (group A), where I participated in conversations with people I knew well, I made no conscious effort to accommodate – as there was no reason for me to do so.

15From a phonological point of view, accommodation towards the BC dialect insofar as segmental and prosodic features were concerned was not difficult for me: being a native to the area, I felt confident that this accommodation was authentic. Grammatically, however, I knew that I ran the risk of accommodating periodically and not systematically, thus giving an impression of over-accommodation and condescension. Table 2 summarises the number, frequency and type of grammatical dialect features that I used in both types of recordings: group A (conversations with second order network contacts) and group B (informal interviews with strangers). It also gives an illustration of each feature. The table shows that I used 7 dialect features in all. All of these were used as variables; i.e. they were sometimes realised in SE and sometimes in BC dialect. The table gives for each group the number of dialect forms over the total number of times the feature occurred. Thus, for example, in group B recordings, 10 out of 15 occurrences of present DO were realised in dialect form (e.g. I dow know), and 5 in SE.

Table 2. Frequency and distribution of the interviewer’s dialect use



Group A

Group B

DO present

I dow know



BE present

typically ay it (SE isn’t it)




I ay seen it



BE past

Where was you born?




past: I come




Her told me




I should hate it to get stolen



16A certain number of observations can be made from the table. Features such as non-inflecting plural forms and preposition changes (e.g. some on ’em for SE some of them), never occurred in my speech, although their SE equivalents were common. The use of the dialect feature SHOULD (i.e. the modal should occurs in contexts where in SE one would expect would, such as in I should hate it to get stolen) occurred in my speech with group A, which would seem to suggest that this feature is part of my natural speech, even when I am not attempting to accommodate towards the dialect. The features HAVE, BE past, COME and HER only occur in samples where I consciously accommodate. The features DO present and BE present occur far more frequently in samples where I am deliberately accommodating, although they are also present in my natural speech. However, in these cases, they are typically I dow know and int it? (tag).

17This quantitative analysis clearly shows that I did indeed accommodate grammatically in group B samples. It remains to be seen in the next section how successful this accommodation turned out to be.

2.2. A qualitative study

18While a quantitative analysis can reveal invaluable information, it is also interesting to examine on a qualitative level certain occurrences of accommodation within Group B samples. As was mentioned above, I only used certain BC dialect forms very sporadically, such as BE past (Where was you born? was they?) as well as BE present (they am). Not surprisingly, these realisations occur almost exclusively at the beginning of the samples, where I was particularly conscious of the importance of making the subject aware of my dialect-speaking status. For example, in extract (1) (dialect forms are highlighted in heavy print), I open the conversation with


Interviewer (L)

It’s a bit chilly in here ay it?

Informant (I)

Yes it is.


It ay very warm.

and at the start of another sample, I ask:



So that’s where you was born then was it?

Other features occur more regularly. In all, I use the tag int it or ay it 13 times in these samples, and I never use the SE equivalent. I frequently use ar as a dialect replacement for yes (10 times in all). For instance:



I’ve gor enough TV at um [home] ay I?


Ar that’s true.

Indeed, in one sample, the informant comments on this use:



Cos I was born here as well.


Was yer?


But I’ve been away for a few years now and I live in France now.




In Strasbourg, just on the German border.


Do yer like it there?


Oh ar it’s great.


Yow’ve still gor a bit o the Black Country though, ‘oh ar’.

However, the samples show that I occasionally responded to a subject’s dialect-form utterance with a SE realisation of the same form:



... her’s always down there.


Is she?

In one sample, I correct my SE in a rather unsuccessful attempt to accommodate. In the preceding section to this extract, I have used SE she 3 times, followed by:



Oh ar that’s it, well she….her worked there for years an’ years an’ years, she worked there when it used to be Holland and Barrett ...

Indeed, the samples show that I occasionally ‘mix’ Standard English and dialect forms in the same utterance in an unnatural way. For instance:



Yes and I used to work there part time.


Did you? What was you doing there then?

19Here the use of the tag did with use to is clearly not dialect (the BC form being Had you?), yet it is immediately followed by a dialect form of BE past (What was you doing).




An they they’re all happy enough I think they’re all pretty happy int they?

20Where, to be consistent, the SE form of BE should certainly have been realised as they’m.

Interestingly, though, the informant replies with a dialect form:



Oh yes they am yes.

21Indeed, this phenomenon occurs several times in the samples: i.e. my unsuccessful attempts at natural dialect being followed by the subject’s immediate dialect realisation of the same feature. For example:



So are your children all living around here still are they, or have they moved ...


Oh they’m very close.

In another sample, a more complicated configuration of dialect plural forms occurs:



It’s bin a good while, above 10 year.


Has it? Have yer bin in a chair for 10 years?


I’ve bin in a chair for 10 year.

22Thus there are several examples in the samples where I unwittingly transform a subject’s dialect form into SE in the very next utterance. There are also several instances where I use unnatural configurations of SE and dialect forms in the same utterance. Yet the overall effect on the subjects’ dialect use seems to be minimal: on all occasions they continue to use dialect forms. Several explanations for this phenomenon can be put forward.

23Firstly, the fact that I alternate between dialect and standard forms is, after all, perfectly natural: all subjects code-switch, to some extent. Thus they may not find my code-switching unusual, even if the exact choice of forms is not always what a typical dialect speaker’s choice would be. Secondly, the examples given above represent only a small percentage of my utterances: most of the time, my dialect use was – I believe – authentic. Thirdly, phonological factors such as intonation and accent also play an important role in successful accommodation, and in this respect I am confident that my speech was authentic. Finally, I believe that several extra-linguistic factors play a very important part in building a rapport between interviewer and subject. As explained in 1.1., I made an effort to mix socially with these subjects before the interviews took place, taking refreshments with them and chatting informally. Moreover, during all of the interviews, I made it clear that I myself used to live locally, and our topics of discussion frequently turned to local topical issues or places. In this way, I hoped to reduce the effect of a ‘stranger’ from outside the community coming to interview them. I am confident that this technique worked. This is perhaps best illustrated in the excerpt below. We had just discovered that the informant lived only a few minutes away from my family. This realisation provoked a sudden and revealing change of topic on her part; once she knew that we belonged to the same community, she immediately began to tell me about an important and personal part of her life:



Oh yes oh do you live round here then?


Well I used to an’ then I went away you see, so ... but me mom and dad still live in Primrose Hill, oh they’ve lived there all their lives.


Oh yes, I might bump into yer one o’ these days. Only I left me husband ages ago, 20 years, never divorced though.

24I believe that the data obtained in these recordings reflect the vernacular style of the informants concerned: I was careful to compare their speech during the interview with their speech during casual in-group discussions (which, for practical reasons, I was unfortunately not able to record), and I noted no discernible difference in the use of the dialect.

25I do believe that in all of these interviews, the fact that I used to live locally played an important part in the authenticity of the resulting data. In my opinion, it far outweighed the occasional unnatural code-switching which may have occurred in my speech. Trudgill (1974) makes the same point in his study of Norwich speech. Being a native of Norwich, his interviews with local inhabitants were made far less difficult by this immediate common identity between himself and the informant. He was also careful to choose a second field-worker who came from the same city. As Edwards, Trudgill and Weltens (1984: 33) note:

26... we feel that this research should be carried out by natives of the area under investigation. [...] The native speaker intuitions of the ‘insider’ often offer valuable insights and short-cuts, and it is certainly easier for a member of a given community to make contact with and record suitable informants without having to worry so much about the sociolinguist’s perennial problem: the observer’s paradox.

3. Accommodation in complex reported speech

27Trudgill (1992: 7) comments that accommodation normally takes place during face-to-face intervention. This corpus contains examples of much more complex types of accommodation in reported speech. Let us examine an example of this phenomenon, taken from a recording of a casual conversation between two working-class friends and neighbours, Betty and Anne, both women being in their sixties, and both BC dialect speakers. The recording lasted around 45 minutes, and contained approximately 8000 words. In extract (13) below, Anne reports to Betty a conversation that she had with her work colleague Jean (not a BC dialect speaker, and Anne’s professional superior). She recounts that Jean had asked her where another work colleague (unnamed) was, and Anne explains to Betty that she had replied to Jean that the unnamed colleague had gone to the toilet. Anne concludes her story by telling Betty that finally, this unnamed colleague came out of the bathroom. Upon close analysis of the extract, it can be seen that when Anne reports her conversation with Jean, she only uses SE forms: she (rather than BC her) as a subject pronoun; wasn’t (rather than BC wor); past tense came (rather than BC come). In the narrative parts of this same short extract, however, Anne uses only BC forms: I says (SE I said); I changed me (SE I changed (myself)); I come out (SE came); her come out then and her went (SE she came out then and she went). Thus Anne clearly indicates to Betty that she had accommodated towards SE when speaking to Jean. Or perhaps it is more accurate to say that Anne chooses to report the conversation in this way – it is, of course, impossible to know if she did indeed accommodate, as it is Anne, the reporter, who has total control of the narrative.

28Furthermore, the shift in dialect also serves to disambiguate the situations of utterance involved, allowing the listener to identify clearly in this extract, for example, the end of the final reported speech segment, indicated by // between came out and so anyway. Dialect forms (and their SE equivalents) have been highlighted in heavy print.

(13)   I went out first an’ I went to the toilet an’ changed me an’ I come out an’ that Jean says // oh where is she Anne, an’ where’s she disappeared to now? // I says // well she must be in the toilet// I says // because she wasn’t like by the sinks when I came out // so anyway her come out then an’ her went.

Another example of the same phenomenon occurs when Anne recounts to Betty a conversation that she had with her friend Margaret, a BC dialect speaker, during which Margaret reports to Anne a conversation that she, Margaret, had had with her doctor.

(14)   Her says // oh no // her says // I day go to work ’cos I wont well enough // her says. //Vick day go to work ’cos he run me up th’hospital, you know // her says // I felt that bad. // Anyway he said he’s got to arrange to see a specialist and it’ll be about 6 weeks and he said // how do you feel about going private // and she says // well how much would it cost // and he says // about 70 pounds. // Yes 70 see I would a thought 30 an’ her says // well her says // it’s a lot o’ money

29The beginning of the narrative includes many BC features: the pronoun her, DO past day, BE past wont, and the clitic determiner th’(hospital). However, as soon as Anne begins to report Margaret who is reporting her conversation with the doctor, there is a clear shift to SE. Phonologically, there is no h-dropping (he’s got to arrange, how do you feel, how much would it cost), which is normally a feature of BC dialect. Moreover, the subject pronoun she replaces the BC form her used at the start of the extract, and the SE plural pounds (rather than BC pound) is used. At the end of the extract, where Anne comments on the situation to Betty, there is a clear return to BC forms (her says well her says it’s a lot o’ money).

30This phonological and grammatical shift towards SE certainly reflects the social importance both Anne and Margaret give to visiting the doctor, who, incidentally, neither speaks the dialect nor has a BC accent. Converging accommodation towards a variety that is seen to be socially desirable seems to be the logical explanation. Again, it is vital to point out that whether Margaret did indeed accommodate in this way is impossible to ascertain; we only have access to Anne’s interpretation of the conversation.

31In conclusion, converging accommodation occurs not only in face-to-face interactions, but also in complex reported speech situations, thus serving to show, in these examples at least, how BC speakers see their interlocutor’s language as socially desirable. It also contributes to disambiguation when reported speech situations contain several embedded levels of discourse.

4. Diverging accommodation or a feminine third person binary politeness distinction?

32As mentioned in the introduction, diverging accommodation, i.e. the distancing of one’s language from that of the interlocutor, in order to disassociate oneself from them, or to assert one’s own identity, or to indicate social distance or disapproval, is a much rarer phenomenon than converging accommodation. It would be very surprising – and rather paradoxical – to find such a phenomenon in the samples studied. However, this section will examine several examples of what initially did seem to be diverging accommodation, before putting forward a more satisfactory explanation involving a feminine third person binary politeness distinction that is specific to the BC dialect.

33At several instances in the data, it was observed that SE forms were used at moments in discourse where the BC forms would have been expected; i.e. informants appeared to code-switch from BC to SE for no apparent reason.

34The following extract illustrates once of these occurrences. Betty asks her friend Anne how Anne’s pregnant daughter-in-law is feeling. Anne uses the BC form her to refer to her daughter-in-law, while Betty, normally a BC speaker, uses the SE form she.


Betty: How is she? Is she alright?

Anne: Her’s all right.

Betty: Is she? She seems to be getting on all right don’t she?

Anne: Her said her had a bad day on Friday.

35One explanation would be that Betty’s BC idiolect does not include the BC form her. Clearly, however, a brief analysis of the rest of the data proves the opposite: Betty uses overwhelmingly, although not exclusively, the BC pronoun in the recorded data.

36It could be, then, that Betty code-switches from BC to SE in mid conversation. In the same utterance, however, Betty uses a BC form don’t (SE: doesn’t) immediately before the SE form she, which seems to invalidate this micro-code switching hypothesis as it would mean that she switches to SE only for the time that it takes to pronounce the pronoun.

37Many other examples of the same phenomenon, always with respect to the choice of pronoun she or her, were found. In extract (16) below, Anne tells Betty about a good friend of hers; Anne uses many BC forms (her, dow (SE: don’t), had used to (SE: did use to)). Betty, too, uses BC forms: she picks up the auxiliary had that Anne has just used (her used to go every day, Barb had) in had she? and she uses a BC past participle sid for SE seen. However, in the same utterance she would seem to code-switch by using the SE pronoun she (Had she?).


Anne: ... her could a probably had a flat or a bungalow down there – her’d a bin near enough because I mean I tell yer this much I dow see Barb as much now, now her used to go every day, Barb had.

Betty: Had she? I’ve never sid her

Anne: Oh yes (unclear) in a car. Her had used to go ... I don’t say her went at weekends, but in the week, and her’d be round there like half past nine...

38Clearly, then, the hypothesis whereby micro-code-switching occurs for diverging accommodation reasons is unsatisfactory. Upon close analysis of the data, another explanation emerges: it would appear that the BC dialect contains two feminine subject pronouns, she and her, and the choice of pronoun depends on the type of relationship the informant has with the referent of the pronoun. (The phenomenon did not extend to the masculine subject pronoun he.) Several clear patterns emerged from the data, suggesting strongly that the variation of she/her depended to a large extent on social and relational concepts which are similar to Brown and Gilman’s (1960) work on the T and V pronoun system in languages such as French and German. In short, her may suggest solidarity, closeness, or, conversely, disrespect, whereas she suggests increased social distance and/or respect. Here, the term ‘increased social distance’ describes, broadly speaking, the case where a particular informant may consider another person to be outside his or her ‘insider group’ of friends, family or work colleagues.

39Thus in examples (15) and (16) cited above, Betty uses she to indicate social distance or respect, showing that she doesn’t know the referents of the pronouns. In (15), the referent is Anne’s daughter-in-law, and in (16), it is Anne’s good friend). Anne, on the other hand, uses her to indicate closeness and solidarity with these two women.

40The data contained two samples that were particularly useful to test this hypothesis. They were the longest ones in the study, each lasting approximately 45 minutes and containing around 8000 words. Over three hundred occurrences of feminine subject pronouns (she or her) were recorded in these two samples. Both had the same situational context: i.e. casual speech between informants who know each other very well. The first was the conversation mentioned in section 3 between Betty and Anne. The second was an informal conversation between (the same) Betty and another friend and neighbour, Pat (also working-class and in her sixties). The informants used the dialect form her as follows (the figures give the percentage of feminine subject pronouns realised as her): Betty (76% in sample (a), 59% in sample (b)); Anne (80%); Pat (46%). In other words, all three women used both forms; for Betty and Anne, the dialect form her was more frequent than she, whereas for Pat both forms were used approximately equally. Close analysis of the data showed that the three women systematically used her when referring to family relations, friends and neighbours that they liked. They used she to refer to the other informants’ family members, to work colleagues of a superior rank, and to other ‘outsider’ group members such as the matron in a nearby old people’s home. They used her to a lesser extent to denote disrespect, but it did occur several times. Negative stance markers such as referring ironically to someone as ‘the millionairess’ as well as my personal knowledge of the relationships involved, made it possible to identify this use. For example, in the first sample, Anne refers to a local teacher who is well-known for her general lack of motivation and enthusiasm, as her, where one would logically expect the pronoun used to be the respectful she:


Anne: Well, Mrs. Scott, the bell, mate, her was off.

41Interestingly, in the same section of the sample, she refers to the headmistress of the school, whom she appreciated, as she.

42The only examples where unexpected pronoun use was found was when there was a general shift towards SE, for reasons that seemed to be linked with topic. Betty, when referring to her granddaughter having taken a music exam, chooses to call her she, and Pat refers to an elderly lady in her mother’s retirement home as she when she explains that she was receiving a visit from the local mayor to celebrate her one hundredth birthday. In both cases, the topic of conversation and the situation suggest contexts in which dialect is not normally spoken: accommodation towards SE forms is thus not particularly surprising.

43Thus what was originally analysed – unsatisfactorily – as diverging accommodation, can be explained more accurately as being due to the BC dialect’s feminine third person binary politeness distinction that is not present in SE.


44This article has attempted to show that accommodation occurs at several levels in the BC data. Firstly, deliberate accommodation towards the informants’ dialect was observed on behalf of the interviewer, in an attempt to reduce the effect of the observer’s paradox when collecting data. A quantitative study of the corpus showed that this accommodation only occurred in samples where the interviewer needed to prove his belonging to the community, i.e. when interviewing people that the interviewer did not know personally. Furthermore, accommodation was partial, from a grammatical point of view: only certain dialect features were employed by the interviewer. A qualitative analysis revealed that accommodation occurred most frequently at the beginning of recordings, and that although it was occasionally inconsistent with so-called ‘natural’ dialect, it did have a positive effect on reducing the observer’s paradox. This, along with the shift of topic towards local community issues, played a significant role in putting the informants at ease and thus producing natural, casual speech.

45Secondly, accommodation towards and away from SE was observed in complex reported speech situations. The reporter switches between BC and SE, using BC when relating past events to her BC speaking interlocutor and using SE in the segments of speech that are reported, where a SE speaker is involved. The move towards SE reflects converging accommodation due to the social prestige given to SE. Thus converging accommodation, either towards BC or towards SE, significantly helped to disambiguate levels of discourse within the dialogue.

46Finally, what initially appeared to be an example of the relatively rare phenomenon of diverging accommodation turned out, in fact, to be primarily due to a binary politeness distinction that is particular to the BC dialect: this dialect uses both BC and SE forms of the feminine subject pronoun, in order to highlight the type of relationship the speaker has with the referent of the pronoun.

Top of page


Brown, Roger and Albert Gilman. “The pronouns of power and solidarity.” In Pier-Paolo Giglioli (ed.), Language and Social Context. Harmondsworth: Penguin, 1960.

Edwards, Viv, Peter Trudgill and Bert Weltens (eds.). The Grammar of English Dialect: A Survey of Research. London: E.S.R.C., 1984.

Gal, Susan. Language Shift: Social Determinants of Linguistic Change in Bilingual Austria. New York: Academic Press, 1979.

Giles, Howard. “Accent Mobility: a model and some data.” Anthropological Linguistics 15 (1973): 87-105.

Heath, Christopher D. The Pronunciation of English in Cannock, Staffordshire. Blackwell: The Philological Society, 1980.

Hernández y Siebold, Nuria. Personal Pronouns in the Dialects of England. A Corpus Study of Grammatical Variation in Spontaneous Speech. 2010. URL: < >.

Higgs, Lyndon. A Description of Grammatical Features and their Variation in the Black Country Dialect. Basel: Schwabe Verlag, 2004.

Labov, William. The Social Stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics, 1966.

Macaulay, Ronald. Locating Dialect in Discourse. Oxford: Oxford University Press, 1991.

Sutcliffe, David. British Black English. Oxford: Basil Blackwell, 1982.

Trudgill, Peter. The Social Differentiation of English in Norwich. London: Cambridge University Press, 1974.

Trudgill, Peter. The Dialects of England. Oxford: Basil Blackwell Ltd., 1990.

Trudgill, Peter. Introducing Language and Society. London: Penguin English Linguistics, 1992.

Wells, John C. Accents of English. Volume 1: An Introduction. London: Cambridge University Press, 1982.

Top of page


1 BATH, STRUT, TRAP and FOOT are part of the lexical sets introduced by Wells (2002) to illustrate clearly and without ambiguity the pronunciation of all vowels in English, in both R.P. and General American.

2 See Higgs (2004) for further information on the grammatical features of the dialect.

Top of page


Electronic reference

Lyndon Higgs, Converging accommodation in the Black Country DialectAnglophonia [Online], 24 | 2017, Online since 28 November 2017, connection on 18 January 2022. URL:; DOI:

Top of page

About the author

Lyndon Higgs

University of Strasbourg, France. LiLPa (EA 1339)

Top of page


Licence Creative Commons
Anglophonia – French Journal of English Linguistics est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Top of page
Search OpenEdition Search

You will be redirected to OpenEdition Search