Introduction

For Samuel Johnson, drafting his Dictionary in the late 1740s, accent was already densely polysemous. It could denote patterns of intonation and the prominence given to certain syllables in pronunciation; antique, he noted, “was formerly pronounced according to the English analogy, with the accent on the first syllable; but now after the French, with the accent on the last” [my emphases]. By poetic license, accent could also signify language or words per se. “How many ages hence| Shall this our lofty scene be acted o’er,| In states unborn, and accents yet unknown”, states Shakespeare’s Julius Caesar in an illustrative citation that Johnson included under this sense. In more general terms, accent, as Johnson confirms, could indicate “the manner of speaking or pronouncing, with regard either to force or elegance”. Supporting evidence from Shakespeare already, however, suggests its potential for qualitative discrimination in this respect, as in the “plain accent” used to describe the forthright speech of Oswald the steward in King Lear or Rosalind’s “finer” accent in As You Like It: “Your accent is something finer than you could purchase in so removed a dwelling.” As Puttenham had indicated in his Arte of English Poesie (1589), reference models for speech are not to be located in the “ill shapen soundes” of craftsmen or carters or, he adds, “others of the inferiour sort”. Even at this point, preference was given to other localized norms, centered on London and surrounding counties within about 40 miles and, in particular, as typified in the usage of educated and courtly speakers –“men ciuill [civil] and graciously behauoured and bred”, as Puttenham affirmed.

As Johnson’s entry for accent suggests, certain meanings are nevertheless prominent only by their absence. Only in the nineteenth century would accent, by a process of synecdoche, come to signify the presence of regional marking in speech per se – so that one might, or indeed might not, in the idioms of English, “have an accent”. “She has a bad figure, she moves ungracefully, perhaps speaks with an accent”, an 1865 citation under accent in the Oxford English Dictionary (OED) confirms. The original definition of accent in OED1, written in 1884 by the phonetician Alexander Ellis, was telling: “This utterance consists mainly in a prevailing quality of tone, or in a peculiar alteration of pitch, but may include mispronunciation of vowels and consonants, misplacing of stress, and misinflection of a sentence. The locality of a speaker is generally clearly marked by this kind of accent.” Illustrative uses include “he has a strong provincial accent” or “an indisputably Irish, Scotch, American … accent”.¹ Citational evidence added in the OED Supplement (1972), here taken from H.G. Wells’s novel The Autocracy of Mr. Parham (1930), confirmed the further consolidation of these ideas. “Underbred contradictory people with accents and most preposterous views”, wrote Wells, providing an unambiguous correlation between “underbreeding” and “accented” speech. Underbred: “Of inferior breeding or upbringing; wanting in polish or refinement; vulgar”, the OED explains. Accent, in Wells’s novel, is made to signal the presence of localized marking alongside assumptions that only those lower in the social spectrum will – or should – possess geographical signifiers of this kind. Other evidence added to the Supplement (now deleted from OED3) made the sociocultural consequences particularly clear: “1956 D. Abercrombie Prob. & Princ. iv. 42: Accent … is a word which, in its popular use, carries a stigma: speaking without an accent is considered preferable to speaking with an accent …. The popular, pejorative, use of the word begs an important question by its assumption that an accent is something which is added to, or in some other way distorts, an accepted norm.”

The location – both social and linguistic – of Abercrombie’s “accepted norm” is equally significant. If “speaking with an accent” had, for Wells, revealed “underbreeding”, the opposite end of the social spectrum lay, as White noted in Words and Their Uses (1881), in “that tone of voice which indicates breeding”. Laden with sociosymbolic values rather different in kind, this form of pronunciation revealed little or nothing of the place of origin of those who used it – whether with reference to what came to be known as “Received Pronunciation” (RP) in Britain, or in the relative homogenization of General American in the United States (see Lippi-Green 1997). As in Abercrombie’s analysis, such speakers, in “popular use”, were regarded as being able to speak “without an accent” at all. George Bernard Shaw’s phonetically-orientated take on the Pygmalion myth in 1914 provides an apt illustration of the sociolinguistic dynamics that can result. Here, the Cockney flower-seller Eliza Doolittle must lose one accent – the geographically marked properties of lower-status London which will, Shaw states, “keep her in the gutter to the end of her days”. Courtesy of intensive phonetic re-education, she instead gains another – an “accentless” RP by which, irrespective of social reality, she will pass for a Duchess at the ambassador’s garden party. Unlike Cockney, which betokened Eliza’s origins – social and regional – in highly specific ways, RP was supra-local, used by speakers “all over the country” as Ellis (1869) had specified, in a speech community characterized by its social meaning as well as its highly restricted membership. As the elocutionist Benjamin Smart (1836) had commented, here with specific reference to accent: “the common standard dialect” is that in which “all marks of a particular place of birth and residence are lost, and nothing appears to indicate any others habits of intercourse than with the well-bred or well-informed, wherever they may be found.” Conversely, it should be remembered that the speech of Northumbrian witnesses, testifying in London in 1861 at the Commission on Mines, was deemed to require an interpreter (Pittock 1997: 118).

While the “received” in other aspects of language practice habitually reflects issues of communality and consensus (see, for example, the early injunction in Cawdrey’s Table Alphabeticall (1604) “to speak as is commonly receiued”), the history of received pronunciation, and its ideologized values, is instead therefore often bound together with the uncommon or nonrepresentative – the language of the privileged few rather than the accented many. The rise of RP as the prime reference accent can, in this light, seem striking. Examining a range of framing discourses such as education, literature, and the mass media, this chapter will explore the changing role and representation of accent, both localized and supra-local, in the history of English. The patterns of endorsement and emulation which are evident in terms of an emergent RP in, say, the eighteenth-century elocution movement or in the prominence of the supra-local in in the training of announcers on the early BBC (Mugglestone 2008) can, for example, stand in recent years alongside evidence of attitudinal resistance, whether in broadcasting or in the accents one might choose to adopt or shed. Here, too, lexical and semantic shifts provide interesting evidence of change. Mockney, a recent entry in OED3 records, is: “An accent and form of speech affected (esp. by a middle-class speaker) in imitation of cockney or of the speech of Londoners; (generally) mockney accent”. As in accounts of the British Chancellor George Osborne’s attempts at linguistic downshifting (in which traditionally stigmatized features are seen as prominent),² a twenty-first century version of Pygmalion might well tell a very different story. “People sneered at the chancellor’s new mockney accent – but it did make him look more human,” wrote Victorian Coren in The Observer in April 2013.

Acts of transformation: the eighteenth-century context

Samuel Johnson, it might be noted, steadfastly retained his Staffordshire accent to the end of his days. This, he declared in 1776, was “the purest English”. Such patterns of local, and linguistic, allegiance offer a useful corrective to habitual readings by which Johnson is often assumed to be single-handedly standardizing the English of his day.³ Yet attitudes to Johnson, and his speech, can in fact usefully illuminate a changing consciousness of accent and pronunciation during this period. David Garrick, the famous actor and theatre-manager, who came to London from Lichfield with Johnson in 1735, followed a very different linguistic trajectory. Some eight years younger than Johnson, it is thanks to Garrick’s mockery of Johnson’s regional marking (a form of speech that Garrick swiftly shed) that we know, for instance, of Johnson’s lengthened Staffordshire vowels in words such as punch. Rather than commendations of Johnson’s accent loyalty, it was perceptions of his “dreadful voice and manner” on which the wife of James Harris, first Earl of Malmesbury (and author of Hermes) likewise comments in April 1775.⁴ Even James Boswell’s Life of Johnson drew attention to Johnson’s “uncouth” tones on their first meeting in 1762 (Pottle 2004: 260): “he speaks with a most uncouth voice”, Boswell wrote in the intended privacy of his London Journal. Of interest too is the diary of Hester Thrale, a close friend of Johnson, who in 1778 decided to award him a score of zero (out of twenty) for “Person and Voice”.⁵

The fact that Thrale decided to initiate an evaluative exercise of this kind among her friends is, of course, also significant in this context. Earlier eighteenth-century comment on differences of speech had been decidedly liberal: “A Country Squire … having the Provincial Accent upon his Tongue, which is neither a Fault, not in his Power to remedy”, Swift had written, for instance, in 1709. “I do not suppose both these Ways of Pronunciation to be equally proper; but both are used … among Persons of Education and Learning in different parts of the Nation”, stated Isaac Watts with similar unconcern (1721: 102). If spelling continued to vary, especially in private use, it clearly also possessed a nationally distributed form; the same was true of the diffusion of a supra-regional grammar. Yet for pronunciation, placed outside the consensus norms of printed texts, there was no public national mode of articulation. The localized, of necessity, remained the norm even if certain modes of pronunciation (e.g., the south-western marking of Somersetshire in Britain) were stereotypically disfavored (see Blank 1996).

The assimilation of accent into regulative discourses of standards and standardization is nevertheless increasingly apparent at this time. Readings of the localized – in the light of what is increasingly promulgated as a supra-regional ideal – can assume strongly negative associations. Boswell himself provides a useful case history. If Boswell is usually remembered in terms of his formative relationship with Johnson, it was in fact Thomas Sheridan, the actor and elocutionist, who was, as Boswell acknowledged, his “Socrates” and mentor. Sheridan’s lectures on elocution – emphasizing, in relation to localized language habits, the importance of a wide-ranging shift in attitudes and practice alike – had prompted Boswell’s immediate enrolment as Sheridan’s private student. “How can consciousness be awoken without information?”, Sheridan had declared (1762: 37): “no man can amend a fault of which he is not conscious; and consciousness cannot exert itself when barred up by habit or vanity”. Boswell proved a most receptive pupil. “Consciousness” led to repeated anxieties about accent, identity, and regional marking. “Mrs. Miller’s abominable Glasgow tongue excruciated me”, Boswell wrote in his London Journal on March 17, 1762 (Pottle 2004: 221). “Habit” was countered by intentionally corrective “information”. Under Sheridan’s instruction, Boswell strove to eradicate all traces (“faults”) of his Scottish origins from his voice. Similar anxieties later led to an assiduous monitoring of his daughter’s speech. If Johnson credited Staffordshire with the “purest English”, Boswell did not agree.⁶

In Sheridan’s rhetoric, images of “received” speech hence exist alongside a determined inculcation of ideas about what should not be “received” at all. Hitherto, he noted (1762: 37), “many provincials have grown old in the capital, without making any change in their original dialect” (a comment it is tempting to read in the light of Johnson’s regionalized speech). In contradistinction, the regional, for Sheridan, is a firm “mark of disgrace”. Placed in the tropes of the “sick” language (an “infection” for which a “cure” is necessary, as Sheridan makes plain), localized speech patterns are framed by the diction of “defect” and “deviation”. The accent proposed as the regulative ideal is rather different – not only in its features but also in the perceptual social and cultural values it is made to suggest. It is “a proof that one has kept good company,” writes Sheridan, “sought after by all, who wish to be considered as fashionable people, or members of the beau monde” (1762: 30). It is, for Sheridan, an indubitable marker of status or social symbol: “Surely every gentleman will think it worth while to take some pains, to get rid of such evident marks of rusticity,” he declares.

Sheridan’s “received” speech is both socially and geographically restricted. Prototypically characterizing upper-status speakers in London, it has, as he continues, hitherto “only [been] acquired by conversing with people in polite life”. Perry (1775) makes a similar point, selecting “the present practice of polite speakers in London” as his intentionally regulative norm. Nevertheless, as a range of writers indicate, a new democratization of access (and of speech) might henceforth be facilitated through education, elocution, and the national power of print. As Sheridan (1762: 30–31) explained:

The difficulties to those who endeavour to cure themselves of a provincial or vicious pronunciation are chiefly three. 1st, The want of knowing exactly where the fault lies. 2ndly, Want of method in removing it, and of due application. 3dly, Want of consciousness of their defects in this point.

As we will see, all three were, in a variety of ways, to be provided as the eighteenth and nineteenth centuries advanced. Whereas Johnson’s Dictionary had merely marked the position of word stress, Sheridan’s Dictionary (1780) had rather different aims. “One main object … is to establish a plain and permanent standard of pronunciation,” the title-page proclaims. Sheridan’s work expounds with striking specificity this shift in “consciousness”, together with the determined positioning of accent within schema of social meaning. It is nevertheless important to see this as part of a wider process. Buchanan’s Linguae Britannicae vera Pronunciatio (1757) was, for example, already starting to explore the provision of an “accurate Pronunciation”, which native speakers as well as foreigners might acquire by means of lexicography. By 1766, Buchanan had published An Essay towards Establishing a Standard for an Elegant and Uniform Pronunciation of the English Language … as practiced by the Most Elegant and Polite speakers. Kenrick’s New Dictionary (1773) likewise promised full information on “Pronunciation … according to the present practice of polished speakers in the Metropolis”. Perry in 1775 made a similar claim. The commodification of accent was also enhanced by the rise of elocution as an industry in a period of marked social change. As an object of desire, the “right accent”, characterized by “elegance” rather than “provinciality”, might also be acquired, as in Sheridan’s teaching of Boswell, or the private lessons offered by a range of other elocutionists across the country (see Benzie 1972).

Pronouncing dictionaries, and other works dedicated to the spoken voice, were disseminated both nationally and internationally,⁷ providing an increasingly detailed and prescriptive reference model. This was /h/-full, possessing the velar nasal /ŋ/ rather than /in/ or /iŋg/ in words such as hopping, /hw/ rather than /w/ in words such as which, using the FOOT-STRUT split, as well as an emergent BATH-TRAP divide. As the elocutionist John Walker (1791: xiii) explained with reference to individual accent modification and the acquisition of “proper pronunciation” (in this instance, the regulative patterning of [v]/ [w]), pronouncing dictionaries were ideally made part of a process of active change:

Let the pupil select from a dictionary, not only all the words that begin with v, but as many as he can of those that have this letter in any other part. Let him be told to bite his under lip while he is sounding the v in those words, and to practice this every day till he pronounces the v properly at first sight: then, and not till then, let him pursue the same method with the w; which he must be directed to pronounce by a putting out of the lips without suffering them to touch the teeth.

Educating accents

“I let other folks talk. I’ve laid by now, and gev up to the young uns. Ask them as have been to school at Tarley; they’ve learnt pernouncing; that’s come up since my day,” comments Mr. Macey in George Eliot’s Silas Marner (1861). As in the localized metathesis of pernouncing, Macey’s speech is made to testify to an earlier educational age. Instruction across the late eighteenth and nineteenth centuries, instead increasingly included spoken alongside written language, with a calculated emphasis on the acquisition of supra-regional markers deemed “standard”. “It ought to be, indispensably, the care of every teacher of English, not to suffer children to pronounce according to the dialect of that place of the country where they were born or reside, if it happens to be vicious,” Buchanan stressed (1757: xli). The potential for social meaning in speech is made particularly explicit: “to avoid a provincial dialect, so unbecoming gentlemen, they are early instructed, while the organs of speech are still flexible, to pronounce properly”, Buchanan persuasively declared. Accent, in private education of this kind, is made a telling object of desire.

“Method”, as Sheridan had explained, was nevertheless vital. The acquisition of regulative (and supra-local) norms depended in part upon “opening a method, whereby all the children of these realms, whether male or female, may be instructed from the first rudiments, in … the art of reading and speaking with propriety and grace” (1762: 225). This process of acquisition was intended to displace existing practice in which habits of pronunciation “depend entirely upon the common mode of utterance in the several places of [children’s] birth and education”. Whether by personal tuition (as for Boswell), educational practice in schools and colleges, or conscious application by the motivated individual, the process – and desirability – of educating accents became a prominent topos. The new genre of the pronouncing dictionary, with its specification of reference models for accent as well as meaning, was presented as particularly useful. The dictionary “must soon be adopted into use by all schools professing to teach English”, wrote Sheridan (1762: 261), a precept also evidently taken on board in the emergent national education system in Britain (see Mugglestone 2007: ch.7). “Rp., received pronunciation”, as Ellis specified, was “that of pronouncing dictionaries and educated people” (1889: 6).

From the point of view of applied linguistics, elocutionary manuals and educational texts provide considerable detail in this respect. Sheridan’s Elements of English (1786), aimed at children from the earliest years, provides an obvious example. This sets out detailed guidance by which a “right pronunciation” is to be acquired – and a “wrong” one displaced. The basis of instruction is phonetic, with the order of instruction being first labials, then dentals, labio-dentals, and “palatines”. Minimal pairs form the basis of exercises and transcriptions offer disambiguation where necessary, as in the recommended distribution of /ʌ/ or /ʊ/ (cut, bull) or /hw/ and /w/ (which/witch) according to supra-regional rather than localized patterns (see, for example, also the specification of rounded [ɒ] after [w] as in want, rather than localized [a]). Only favored variants are recorded.

Evidence of the implementation of instruction of this kind is particularly important. Poole’s The Village School Improved (which had three editions 1813–1815) offers considerable detail of the ways in which, in Enmore in Somerset, children were encouraged to abandon “provincial” forms in favor of supra-local models. Reading aloud became an exercise in discrimination. “Even a coarse or provincial way of pronouncing a word, though sanctioned by the general practice of the district, is immediately noted by the teacher; and exposes the child … as much to the correction of those below him, and consequently to the loss of his place, as any other impropriety in reading would do” (Poole 1815: 40–41). The hierarchical ranking of the class is particularly telling, offering a microcosm of the kind of top-down models of convergence that contemporary works on elocution advocated. Local children, Poole admitted, have habitually “heard and spoken a broad provincial dialect”. Learning “to pronounce with propriety” could be challenging: “The more remote the dialect of the [child’s] country is from the propriety of the language, the greater is the embarrassment experienced … when he begins to be instructed according to the new and improved system” (1815: 41). Nevertheless, the benefits are presented as incalculable: “this embarrassment is merely temporary” but “permanent advantages are sure to follow”, not least in the “intelligent, discriminating manner of reading” and “purity of pronunciation” that will, in the end, be acquired.

Teaching manuals from later in the century provide further evidence of the ways in which reference models of accent were incorporated within general educational practice and assessment. Morrison’s Manual of School Management, which went through three editions (1859–1863), presents a useful example. Originally “designed for the use of students attending the Glasgow Free Church Training College”, the manual sets out recommended methods of instruction on the basis of tried and tested methods. “Nothing has been set down which experience has not proved attainable,” Morrison stresses (1863: iii). Exercises within individual chapters are given as aligned with the Committee of Council of Education “with the view of directing attention to the points considered important by the Inspectors of Schools”. An extensive section details “the correct use of letters, the signs of sounds”. For the teacher, “the first thing to be done is to analyze the language into its simple elementary sounds”; these again include contrastive medial vowels in cut and bull, cat and cast, as well as use of the velar nasal /ŋ/ in words such as skipping. As in Sheridan, minimal pairs are advised to enable facility in reading and speaking alike. A section headed “Correct Pronunciation” outlines the principles by which the teaching of reading includes not only comprehension but articulation in the prescribed way: “the first essential requisite in good reading is correct pronunciation” (1863: 125). This, Morrison (1863: 125) points out, is dependent on the teacher suppressing (a) his/ her own regional marking and (b) those of the children in his/her care:

There is no security that the pupils acquire correct pronunciation, unless the teacher be able to give the example. Accordingly the teacher who is anxious to be in this, as in all things, a model, should strive during his preparatory training to acquire a thorough knowledge of English pronunciation. This can only be done by careful observation of good speakers, or, if need be, by a course of lessons with an accomplished and trust-worthy teacher. Whenever the young teacher hears a good speaker pronounce a word differently from what he has been accustomed to, he ought to note it, and never rest satisfied until he has ascertained the correct pronunciation. He will be amazed at the benefit such a course will confer. (1863: 126)

While the teacher’s acquisition of “correct orthoepy” is made central to teaching ability in this context, Sheridan’s earlier emphasis on “method” is also clear. “The only effectual method by which [the teacher] can secure good pronunciation among his pupils, is to insist that they pronounce every word correctly,” writes Morrison: “Constant correction … will alone accomplish the desired result.” An educated accent is specified as one devoid of the “peculiarities of pronunciation” which characterize “various districts”, whether in terms of “a constant tendency to shorten the long vowels” or “in others to lengthen the short ones”, or in the presence other regionally marked features (1863: 126). The normative remit of the teacher is evident: “we advise the teacher, whenever he finds himself located in a particular parish, to observe carefully the prevalent peculiarities; and, when he has done so, vigorously to set himself to correct them among his pupils” (1863: 127). Education reveals, in essence, the firm institutionalization of an ideology in which pronunciation can be divided on standard/subordinate models.

Morrison’s strictures are paralleled in a range of other teaching manuals, as well as in school inspectors’ reports where articulation (and the absence of regional marking) is often presented as proof of educational success. Recitation – the reading out of a passage with “proper” elocution – was a popular aspect of assessment in which the presence of regional markers could be viewed as testimony not only to local identity but, as other educationalists admonished, as indicators of “Defective Intelligence” per se. It was in these terms that John Gill, one of the most influential writers of teaching manuals in this context (see Hole 2003) chose to orientate his discussion of features such as zero-realization of /h/ or the nonuse of /ʌ/ in cut. The classification of purely phonetic features under “Defective Intelligence” amply confirms the negative repercussions of applied language attitudes in educational practice of this kind.

Self-education presents a further domain in which attitudinal shifts to regionally marked speech, and the attempted inculcation of a supra-local model, is in evidence. Texts on pronunciation and elocution often recommended assiduous self-application. It is, however, specific evidence on individual receptiveness to such dictates that can be most illuminating. Prescriptive rhetoric provides merely one side of the story. A useful snapshot here is provided by Michael Faraday, the scientist (and famous lecturer) who began life as the son of a blacksmith in working-class London. It was in this context of self-improvement that Faraday’s interest in language, and specifically pronunciation, began. By 1813, he had established, with other members of the local City Philosophical Society, a “mutual improvement plan” whereby some half a dozen friends met “to read together, correct, and improve each other’s pronunciation” (see Mugglestone 2011). Five years later, this plan was extended by Faraday’s decision to attend Benjamin Smart’s lectures on elocution, from which Faraday’s detailed notes, running to some 150 pages, remain in the Royal Institution archives in London.

Faraday noted, in full, Smart’s maxim: “Always pronounce words according to the best usage of the time … defects or provincialities must be corrected by a dictionary for which purpose I would recommend Walker’s or by reference to those who are already correct.” Comments on “defective articulation”, and its needful remedy, receive equal attention: “H is … the most subject to a corrupt pronunciation and therefore requiring our early attention,” Faraday’s notebook records; “The person should practice … lists of words beginning with H, then in mixed lists of words some beginning with H, and some with a vowel and lastly with the introduction of the words commencing with H mute.” As Smart pointed out, lectures should be accompanied by active practice, not merely passive listening. “Man”, Smart added (in another maxim noted down word for word), “is an improving animal … that man only is to be condemned and despised, who is not in a state of transition. We are by our nature progressive.” Like Sheridan for Boswell, Smart was Faraday’s phonetic mentor, in a connection that lasted until the 1850s.

Attitudes, accent, and popular culture

Popular culture also acts as a domain in which the information central to Sheridan’s recommended shift in “consciousness” can come into play. The shifts in language practice attested by Boswell and Sheridan, for instance, testify to that process of enregisterment – a cultural awareness of a set of social meanings associated with specific varieties of speech as detailed by Agha (2003, 2005). Cockney, Scots, as well as speech varieties that participate in what Lippi-Green describes as “the myth of non-accent” (1997: 41) all exist, among other varieties, as enregistered forms across the nineteenth century – and, as Shaw’s Pygmalion affirms, into the twentieth century too. Literary texts, and the conventions of representation they adopt, can reflect and foster perceptual meanings in this respect with ease.

As in the following extract from George Gissing’s Born in Exile (1892), conventional orthographical patterning is placed in contrastive distribution with strategic patterns of respelling in the representation of direct speech. Text conventions of this kind rely on acts of reception by which unmodified spelling will, by implication, suggest the standard proprieties of “educated” speech. A social as well as linguistic divide is made to separate Godwin Peake, a student at Whitelaw College, and his uncle; here, a range of approximations denotes the urban vernacular of the London underclass that Godwin’s Uncle Joey retains. The textual as well as social asymmetries in representation intentionally encode divisions of identity, education, and status. Yet, as Blake (Austin and Jones 2002: xvii) warns, “Any spelling which differs from th[e] standard may seem bizarre because it is strange; and what is bizarre may often seem ludicrous or comic.” Visual disparities of form readily reinforce normative readings of one variety against what can be made to seem unambiguous infelicities and errors in another. Here, stigmatized features such as [Ø] for [h] in ‘ow (how), or [in] rather than [iŋ] (caterin’ against catering) are signaled by the inserted apostrophe. As a graphemic marker, this engages with models of deficit rather than difference (indicating the absence of something that “should” be there). Other features (the absence of sandhi phenomena in a openin’, a ‘int) are reinforced in intentionally negative readings by their co-occurrence with nonstandard grammar (e.g., as relative in “give a ‘int to the young gents as you might come”, alongside multiple negation). The use of socially disfavored lexical items is equally marked. Gent, as OED1 specified in 1899, was “only vulgar, exc. as applied derisively to men of the vulgar and pretentious class who are supposed to use the word, and as used in tradesmen's notices”.

'This ain't no wye of caterin' for young gents at Collige!' he exclaimed. 'If there ain't a openin' 'ere, then I never see one. Godwin, bo-oy, 'ow much longer'll it be before you're out of you're time over there?'

'It's uncertain – I can't say.'

'But ain't it understood as you stay till you've passed the top standard, or whatever it's called?'

'I really haven't made up my mind what to do.'

'But you'll be studyin' 'ere for another twelve months, I dessay?'

'Why do you ask?'

'Why? cos s'posin' I got 'old o' this 'ere little shop, or another like it close by, me an' you might come to an understandin'—see? It might be worth your while to give a 'int to the young gents as you're in with—eh?'

Godwin was endeavouring to masticate a piece of toast, but it turned to sawdust upon his palate.

Even where pronunciation features are likely to be shared by speakers of different social identities (as in weak forms such as of in positions of low stress, or the patterning of ellipsis), they are typically allocated as “accented” and, by implication, “nonstandard” features. Such skewed patterns of representation heighten the assumed contrast between a “standard” – and unmarked – supra-local discourse, against other varieties that are marked, socially and regionally, in a range of ways (see also, for example, American novels and the contrastive marking of accents of the South). Textual patterning of this kind was, by the end of the nineteenth century, a widespread feature of canonical and noncanonical texts alike, appearing in popular journals, newspapers, and magazines, as well as novels.

Factual works can, in fact, be equally productive in the level of language consciousness that they reveal. Entries in the first edition of the Dictionary of National Biography (Stephen and Lee 1885–1891) present particularly useful examples, frequently drawing attention to accent as a salient property of identity. “So perfectly fitted was Ainley, both in looks and voice – from which the north country accent had gone during his training under Benson – that he became famous on the first night,” we are informed of the actor Henry Ainley; “His short, stout appearance and strong northern Irish accent did not endear him to his contemporaries; Disraeli remarked ‘What is that?’ on first hearing Biggar speak in the house,” the entry for the politician Joseph Biggar states. Entries for Frederick Alexander (“His cultured voice had no trace of regional accent”) or Sir Francis Beaufort (“rejected by a school in Cheltenham on the ground that his Irish accent would corrupt the speech of the other boys”) share an emphasis on pronunciation as a reference point for social identity. The fact that, in the relatively brief accounts provided, it was seen as important to confirm that William Huskisson had “a most vulgar uneducated accent” or the politician John Felden had a “strong provincial accent” likewise attests to the perceived salience of attitudes of this kind. The DNB1 entry for the actor Hannah Brand, and the sense of unacceptability her regional accent elicited, is particularly interesting in the light of shifts in language ideology (and recommended changes in praxis) at this time: “Two years later, on 20 March 1794, Brand appeared at the York theatre, playing Lady Townly in Vanbrugh's The Provoked Husband. Her manager there, Tate Wilkinson, complained of her old-fashioned dress, provincial accent, conceit, and contradictory passions. All of these provoked the audience, and her performance “met with rude marks of disgustful behaviour”.

The broadcast voice

Brand’s castigation in terms of accent was intensified because of her prominent position upon the stage – an early model of a broadcast voice. Broadcasting in its modern sense is, of course, a much later phenomenon. In Britain the British Broadcasting Corporation (BBC) – originally the British Broadcasting Company – instituted national radio broadcasting in 1923. Its remit, as its Director General, John Reith, stressed, was that of public service broadcasting. Ameliorative and beneficial, it was to provide opportunities for access to high culture in what an article in The Observer on 18 July 1926 described as a “University of taste”. Language was seen as another aspect of such remedial change: “Wireless … can do much to repair … one of the most conspicuous failures of elementary education in raising the quality of common speech.” The Observer continued: “It could establish – in time – a standard voice analogous to the ‘standard yard’ and the ‘standard pound”’ (“Pronunciation Problems” 1926: 17). As Cecil Lewis, an early employee at the BBC, confirmed (1924), “it has often been remarked – and this is one of the responsibilities that are indeed heavy to carry – that the announcing voice sets a fashion in speaking to thousands of homes and should therefore be faultlessly accurate.” The ideal, Lewis added, was that of “accentless” speech.

Reith was particularly engaged with the idea of broadcast English as a reference model. Elaborated in his Broadcast over Britain (1924), this led to increasingly stringent policies on the kind of accents deemed suitable for announcers. “We are daily establishing in the minds of the public the idea of what correct speech should be and this is an important responsibility,” a BBC directive of 1925 specified. As for Sheridan, images of top-down convergence and the need for corresponding emulatory endeavor are marked. As The Guardian wrote in December 1932, the BBC’s agenda seemed to be that of “levelling up” pronunciation. “You cannot raise social standards without raising speech standards,” Arthur Lloyd James, responsible for the training of announcers on the early BBC, had declared. As The Guardian reported, “The case for such attempts to level up pronunciation, as put by Mr. Lloyd James, is that it is the business of State education to remove improper, or at any rate socially unpopular, forms of speech behaviour, because this is in practice an obstacle to getting on in the world.”⁸ If the BBC was, in this, responsive to pre-existing language attitudes, a clearly interventionist remit was also assumed, as Lloyd James (1927) indicates:

For some reason a man is judged in this country by his language, with the result that there is, broadly speaking, a sort of English that is current among the educated and cultured classes all over the country. It has little local variations, but these are of no matter, and a man who has this sort of accent moves among the rest of his fellow country men without adverse criticism.

This type of speech avoids the lapses of the uneducated and the affectation of the insufficiently educated at both ends of the social scale, and it is the duty of the BBC to provide this sort of speech as often as possible.

While regional speech appeared on local broadcasting, the early BBC effortlessly inculcated the sense of a supra-regional accent as one of its quintessential features, reinforced through accent training in which RP’s hegemony was indubitable. That the same practices extended to Australia and Canada (Price 2008), where RP also came to dominate in news broadcasting and announcing, is still more striking.

Belief and behavior: convergence and divergence

Received English, and the acts of reception that surround it, can nevertheless be more complex than the elocutionary rhetoric of Sheridan, Buchanan, or the early BBC can suggest. If responsibility is overtly assumed for the dissemination of one particular “standard” model through the “noble art of printing” by Sheridan or by direct transmission of particular accents (and their associative meanings) on the early BBC, the reality of language practice can, of course, continue to be conspicuously diverse. A supra-regional mode of speech (as Ellis already indicated in the late nineteenth century), RP spans a spectrum of related forms and emerging/obsolescent variants; yod-presence exists alongside yod-absence in words such as suit in Ellis’s transcribed forms, just as monophthongal variants existed alongside diphthongs in words such as mate. Perry’s ambition to fix a social model of speech has, in this respect, failed. In Britain, RP is today used by a minority – usually estimated at between 3 and 5% of the population (see, for example, Hughes, Trudgill, and Watt 2012).

Well over 90% of the population has, in these terms, maintained some degree of localized marking in their speech. Accent as a social symbol hence testifies to far more than the indices of the “well-bred”, as stressed by Smart, or familiarity with “good company” as Sheridan proclaimed. Outside those accents promoted as “educated” stand, for example, the authority of vernacular culture, of accent loyalty, and of resistance to the ideological hegemonies in which one type of accent alone is favored and the others proscribed. Reactions to the early BBC, and the acts of speech standardization that it attempted to foster, are particularly useful in this context. The privileging of particular forms of speech on the airwaves was not necessarily without resistance. As The Manchester Guardian stressed in 1927, “In self-expression we are heretics all, proud of our dialects and our difference.” Acknowledging that “the B.B.C. … has attempted to achieve a pact of pronunciation within these islands”, it queried whether this could or should be made a shared norm for all. After all, here against the rhetoric of the “accentless”, forms of this kind were profoundly “accented” when seen from, say, the perspective of speakers in the Midlands and the North. If RP was supra-regional in use, it remained distinctly southern in its patterning of words such as fast and bath, cut and bull. Attempted standardization, the writer continued, was “in many respects a surrender to the slovenly and drawling speech of the Southern English and will be promptly disregarded by all self-respecting speakers of the language” (“Speech control”, 1927: 8). Normative readings of accent varieties are not always shared. Images of “disgrace”, in Sheridan’s terms, can be countered by those of pretension. As in Gaskell’s Mary Barton (1848), the question of who precisely “talks the right language” can already be made depending on where you are coming from: “‘’You're frightening them horses,’ says he, in his mincing way (for Londoners are mostly all tongue-tied, and can't say their a's and i's properly)”, as the Manchester-born John Barton is made to aver.

“Is it wrong for a person to change their accent?”, The Observer in April 2013 demanded. The social rhetoric it explored exposed the wide-ranging assumptions that have, since the eighteenth century, often informed popular writing on accent. Question of class, social, prejudice, and discrimination all surface in such debates. Since no one accent is inherently better, the arbitrariness of attributions of “disgrace” and “polish” is all too clear. Sheridan’s intended democratization in terms of accent now firmly rests in the shared understanding of the perceptual nature of varieties, rather than in pressures for conformity to a top-down ideal. Prestige, too, in this light, is multidimensional. Covert and overt prestige do not pull in the same direction (see, for example, Watson 2006). Specified norms can be rejected; RP, rightly, has been displaced in Australian broadcasting (as well in other domains where national varieties of English now assume pride of place). Like other varieties once promoted as inviolably “correct” (see Lippi-Green 1997), RP is now understood as profoundly accented, not only in its phonological patterning but in the social meanings it has traditionally assumed. Even in news broadcasting on the BBC, it has largely lost its dominance, while transcription policies in OED3 likewise reflect a commitment to varietal forms. The revised entries of the new DNB (Matthew, Harrison, and Goldman 2004) are likewise substantially different in emphasis and orientation. If the supra-local remains a model in language teaching, the hyperlectal features of U-RP (upper-class RP) are not advocated, while notions of the “received” can prompt evident unease. “Because of the dated – and to some people objectionable – social connotations, we shall not normally use the label RP (except consciously to refer to the upper-class speech of the twentieth century),” write Collins and Mees (2003: 3–4). Such shifts of social symbolism are interesting. Alongside the disfavoring of U-RP is, as Coupland and Bishop (2007) confirm, a clear valorization of speakers’ own varieties in many (but not all) cases, alongside a decreased responsiveness to supra-local norms in younger speakers. The sociophonetic landscape can nevertheless remain complex. Even in 2013, issues of regional accent and educational delegitimization can still recur. “Cumbrian teacher told to tone down accent,” as The Independent newspaper stated in November 2013, reporting the views of education inspectors on a school in Berkshire. Alongside the rise of mockney and the incorporation of once-stigmatized features such as glottalization within modern RP, the perceptual legacies of the past can linger on.⁹

REFERENCES

1. Agha, A. 2003. The social life of cultural value. Language and Communication 23: 231–273.
2. Agha, A. 2005. Voice, footing, enregisterment. Journal of Linguistic Anthropology 15: 38-59.
3. Austin, F. and Jones, B. 2002. The Language and Craft of William Barnes, English Poet and Philologist (1801–1886), Introduction by Norman Blake, Lewiston, NY; Lampeter: Edwin Mellen Press.
4. Balderston, K. (ed). 1942. Thraliana, the Diary of Mrs. Hester Lynch Thrale (later Mrs. Piozzi) 1776–1809, 2 vols., Oxford: Clarendon Press.
5. Benzie, W. 1972. The Dublin Orator: Thomas Sheridan’s Influence on Eighteenth-Century Rhetoric and Belles Lettres, Leeds: University of Leeds.
6. Blank, P. 1996. Broken English: Dialects and Politics of Language in Renaissance Writings, New York: Routledge.
7. Boswell, James. Boswell’s Life of Johnson; Together with Boswell’s Journal of a Tour to the Hebrides and Johnson’s Diary of a Journey into North Wales, 6 vols., George Birkbeck Hill (ed.), revised and enlarged by L.F. Powell, 1971, 2nd edition, Oxford: Clarendon Press.
8. Buchanan, J. 1757. Linguae Britannicae vera Pronunciatio: or, a New English Dictionary, London: A. Millar.
9. Buchanan, J. 1766. An Essay Towards Establishing a Standard for an Elegant and Uniform Pronunciation of the English Language, London: Edward and Charles Dilly.
10. Cawdrey, Robert. 1604. A Table Alphabeticall Contayning and Teaching the True Writing and Vnderstanding of Hard Vsuall English Words. London: Edmund Weauer.
11. Collins, B. and Mees, I. 2003. Practical Phonetics and Phonology. A Resource Book for Students, London: Routledge.
12. Coren, V. 2013. “George Osborne, gawd bless yer”, The Observer, 7 April.
13. Coupland, N. and Bishop, H. 2007. Ideologised values for British accents. Journal of Sociolinguistics 11: 74–93.
14. “Cumbrian teacher told to tone down accent”. 2013. The Independent, 19 November.
15. Ellis, A. 1869–1889. On Early English Pronunciation, 5 vols., London: Trübner.
16. Eliot, G. (1861). Silas Marner: The Weaver of Raveloe, Edinburgh and London: William Blackwood and Sons.
17. Gaskell, E. 1848. Mary Barton. A Tale of Manchester Life, 2 vols., London: Chapman and Hall.
18. Gissing, G. 1892. Born in Exile, London and Edinburgh: Adam and Charles Black.
19. Hole, N. 2003. Nineteenth-century method manuals with special reference to English literature. In Paradigm, F. Austin and C. Stray (eds.), The Teaching of English in the Eighteenth and Nineteenth Centuries. Essays for Ian Michael on his 88th Birthday: 68–75.
20. Hughes, A., Trudgill, P., and Watt, D. 2012. English Accents and Dialects. An Introduction to Social and Regional Varieties of English in the British Isles, 5th edition, London: Edwin Arnold.
21. Johnson, S. 1755. A Dictionary of the English Language, 2 vols., London: W. Strahan.
22. Kenrick, W. 1773. A New Dictionary of the English Language, London: John and Francis Rivington.
23. Lass, R. 2000. The Cambridge History of the English Language. III. 1476–1776. Cambridge: Cambridge University Press.
24. Lewis, C. 1924. Broadcasting from Within, London: G. Newnes.
25. Lippi-Green, R. 1997. English with an Accent: Language, Ideology, and Discrimination in the United States, London: Routledge.
26. Lloyd James, A. 1927. “A B.B.C. expert on pronunciation”, Teaching World, 21 September.
27. Masters, S. 2013. “George Osborne’s ‘Man of the People’ accent ridiculed”, The Independent 26/6/2013. http://www.independent.co.uk/news/uk/politics/george-osbornes-man-of-the-people-accent-ridiculed-8675419.html.
28. Matthew, H., Harrison, B., and Goldman, L. 2004. Oxford Dictionary of National Biography, Oxford: Oxford University Press. http://www.oxforddnb.com/.
29. Morrison, T. 1863. Manual of School Management, 3rd edition, Glasgow: William Hamilton.
30. Mugglestone, L. 2003. Sheridan in the Schoolroom. In Paradigm, F. Austin and C. Stray (eds.), The Teaching of English in the Eighteenth and Nineteenth Centuries. Essays for Ian Michael on his 88th Birthday: 22–28.
31. Mugglestone, L. 2007. Talking Proper. The Rise and Fall of the English Accent as Social Symbol, Oxford: Oxford University Press.
32. Mugglestone, L. 2008. BBC English. In the beginning. In Broadcast English, J. Schwyter, D. Maillat, and C. Mair (eds.), Arbeiten aus Anglistik und Amerikanistik 33, 197–215, Tübingen: G. Narr.
33. Mugglestone, L. 2011. Benjamin Smart and Michael Faraday: The principles and practice of talking proper in nineteenth-century England. In Contours of English and English Language Studies: In Honor of Richard W. Bailey, M. Adams and A. Curzan (eds.), 87–107, Ann Arbor: University of Michigan Press.
34. “Our London Correspondence”. 1932. The Manchester Guardian, 15 December: 8
35. Perry, W. 1775. The Royal Standard English Dictionary, Edinburgh: David Wilson.
36. Pittock, M. (1997). Inventing and Resisting Britain: Cultural Identities in Britain and Ireland, 1685–1789, Basingstoke: Macmillan.
37. Poole, J. 1815. The Village School Improved. 3rd edn. Oxford: Oxford University Press.
38. Pottle, F. (ed.). 2004. Boswell’s London Journal, 1762–1763, 2nd edition, Yale: Yale University Press.
39. Price, J. (2008). New news old news: a sociophonetic study of spoken Australian English in news broadcast speech. In Broadcast English, J. Schwyter, D. Maillat, and C. Mair (eds.), Arbeiten aus Anglistik und Amerikanistik 33, 285–310, Tübingen: G. Narr.
40. “Pronunciation problems. Further list by the B.B.C.” 1926. The Observer, 18 July: 17.
41. Puttenham, G. (1589). The Arte of English Poesie, London: Richard Field.
42. Reith, J. 1924. Broadcast over Britain, London: Hodder and Stoughton.
43. Shaw, B. 1914. Pygmalion: A Romance in Five Acts, London: Constable.
44. Sheridan, T. 1761. A Dissertation on the Causes of the Difficulties, which Occur in Learning the English Tongue, London: R. and J. Dodsley.
45. Sheridan, T. 1762. A Course of Lectures on Elocution, London: J. Dodsley.
46. Sheridan, T. 1780. A General Dictionary of the English Language, London: J. Dodsley.
47. Sheridan, T. 1786. Elements of English, London: C. Dilly.
48. Smart, B. 1836. Walker Remodelled. A New Critical Pronouncing Dictionary of the English Language, London: T. Cadell.
49. “Speech control” (1927). The Manchester Guardian, 17 June: 8.
50. Stephen, L. and Lee, S. (eds.) 1885–1891. Dictionary of National Biography, London: Smith, Elder, & Co.
51. Swift, J. 1709 A Project for the Advancement of Religion, and the Reformation of Manners.
52. Watts, I. 1721. The Art of Reading and Writing English, London: John Clark.
53. Walker, J. 1791. A Critical Pronouncing Dictionary and Expositor of the English Language, London: G. Robinson.
54. Watson, K. 2006. Phonological resistance and innovation in the North-West of England. English Today 22: 55–61.
55. Wells, H.G. 1930. The Autocracy of Mr. Parham, New York: Doubleday, Doran and Co.
56. White, R. (1881). Words and Their Uses, Past and Present, New York: Sheldon.