Pronunciation model: Bermudian English

View the key for Bermudian English here.

Bermuda is an island territory in the North Atlantic, situated 650 miles from the coast of North Carolina, USA. It is a British Overseas Territory spanning 21 square miles inhabited by a population of an estimated 65,000-72,000 (though many of its 181 islands, islets, and rocks are uninhabited). Bermuda was first inhabited when a British ship was wrecked on its shores en route to the colony of Jamestown, Virginia, in 1609, and subsequently settled by the English with settlers arriving from 1612 onwards.

Slavery has played a significant role in the evolution of Bermuda. Slavery was established very early in its settlement, with most coming from the Caribbean and North America rather than Africa, which Hall (2018) notes is consistent with Bermuda’s lack of a creole, as is the nature of Bermudian slavery in terms of the enslaved working closely with owners on small homesteads (rather than larger groups working on large estates as found in North America and parts of the Caribbean). Most enslaved Black and Native American people arrived in the first 50-60 years of Bermuda’s history, with subsequent steps taken to limit the Black population as it grew. Slavery being abolished in 1834, immigration of Portuguese workers was subsequently incentivized and such immigration continued to grow, even accelerating in the 1920s. The early 1900s also saw West Indian construction workers brought to Bermuda, while the USA maintained naval and air bases in Bermuda from 1941-1995. The attractiveness of Bermuda to international businesses has seen the arrival of relatively wealthy immigrants from the UK, USA, and Canada from the first half of the 20th century to the modern day.

Ethnically, then, Bermuda reflects this complex history (the 2016 census indicated 52% Black, 31% white, 9% mixed race, 4% Asian, 4% ‘other’). Hall (2018) highlights that the Portuguese-descended white community in Bermuda is very much under-studied, but that the Black and white populations in Bermuda are disparate in economic, political, and cultural terms. Linguistically, Bermuda’s dialect is also under-researched, but there is a uniqueness to the form of English that has developed on Bermuda, since there was no existing language in the territory for it to interact with and it falls between a colonial and postcolonial status without being either. Dialect contact rather than language contact is likely to have been the main influence. As Hall explains: ‘[…] we can identify a number of varieties of English likely to have contributed to the development of [Bermudian English]. In early Bermuda, these are London English, varieties of English from the southern ports of England, and developing Caribbean English varieties, particularly from the Bahamas and the Turks and Caicos Islands. From the mid-17th century, speakers of Irish and Scottish Englishes arrived in Bermuda […] In the late 19th and early 20th centuries, the arrival of Caribbean and Azorean Portuguese migrants introduced further input. And over the last century, large numbers of American, Canadian and British English speakers have settled in Bermuda.’ (2018, p.34). There is an important historical and geographical relationship with the southeast coast of the USA, particularly Virginia and the Carolinas.

The common presentation of Bermudian English as reflective of 17th/early-18th century British English (i.e. that which was exported with US settlers) is understandable given a cursory glance, but overstated given the complexity of influences. It would also be overstating to assume just one form of Bermudian English exists, since as Hall notes, ‘it is likely that Black Bermudian English is more influenced by Caribbean and Portuguese input than White Bermudian English, and, like African American English, it is likely to have diverged from white varieties as a result of segregation’ (2018, p.38). Trudgill (2019) notes that while African-Bermudian English shares features with Caribbean English, he asserts that the speech of basilectal white speakers should be grouped with varieties of North American speech, supporting the description of Ayres (1933) in this regard.

Hall (2019) suggests that White Bermudian English and Black Bermudian English may best be considered ‘overlapping subvarieties’ of Bermudian English, with some significant differences. The OED Bermudian English (BerE) model presented here does not distinguish between these, but attempts to converge on their similarities, grounded in Hall’s Black Bermudian English (BBerE) description with features of traditional basilectal white Bermudian English (TBWBerE) described by Trudgill (2019) and Ayres (1933).

iː, ɪjə
æ, e
e, ejə
ɛə, ejə
e, ɛjə
ɔː, ɑ
ɑːɪ, ɒɪ
CLOTHɔə, ɔː, ʊəNORTHɔə(r), ɔː(r)VOICEɔɪPIANOi
STRUTɒ, ʌFORCEɔə(r), ɔː(r), ʊəLOUD/NOW
əː, ɵː
əʊ, ɵː

Many vowels in BerE are contextually conditioned, with one form before voiced consonants in monosyllabic words and stressed final syllables of polysyllabic words (hereafter VC) and another elsewhere (e.g. same-syllable voiceless consonants, nonfinal syllables of polysyllabic words; hereafter Vc). Some of these exhibit ‘southern Breaking’, a term employed by Wells (1982) to describe a pattern of some southern US English monophthongs having diphthongal variants in certain contexts. In BerE, KIT is Vc /ɪ/, VC /ɪjə/, but note the salient variant /iː/ before /ɡ, ŋ/ in Black Bermudian English (BBerE).

The DRESS-TRAP merger reported for BBerE can also have a Vc-VC distinction (same for the DRESS and TRAP vowels for TBWBerE) but are most commonly monophthongs. There is also a pre-/ɡ, ŋ/ BBerE pattern of /e/. LOT and FLEECE also have a Vc-VC distinction for at least one of the two varieties. Since these tendencies typically seem to apply only where the voiced consonant within the same syllable, a word such as product has a first vowel of /ɑ/, rather than the /ɔː/ that may be expected with prod. PRICE, MOUTH, and GOAT are often different in pre-voiceless-consonant environments from their pre-voiced-consonant counterparts (i.e. these are the qualities before same-syllable voiceless consonants). However, in open syllables, PRY and NOW follow PRIDE and LOUD, but GO follows GOAT rather than ROAD. STRUT’s rounding is quite minimal and a less common /ʌ/ variant is also found, while GOOSE is fronted and can even reach a [yː] quality (the BBerE pre-/l/ [ɨə] quality is not reflected here). FOOT appears to have a diphthongal [ʉə] quality before /d/, but data is sparse as to whether this quality applies more widely and for simplicity is not reflected here.

Two distinct forms of BATH are found: older BBerE speakers use /ɑː/ while TBWBerE and younger BBerE speakers use /aː/.

The increasing rhoticity of the BBerE is starting to affect the START, NORTH, and FORCE vowels, with NORTH/FORCE predicted to eventually unmerge from CLOTH/THOUGHT. BBerE’s NURSE-NEAR-SQUARE merger to a rhotic mid central vowel is not mirrored in TBWBerE (i.e. no NEAR-SQUARE merger).

The variability of the CURE vowel is significant, but it is almost always non-rhotic. TBWBerE has a rounded central monophthong for GOAT. VOICE has several possible BBerE qualities, but most commonly in casual speech is best reflected as /ɔɪ/.

The rhotacization of vowels in BBerE does not seem to have yet affected schwa in LETTER.

Consonantally, the most salient feature is the use of /β/ for words which in British or US English would have /v/ or /w/, reflecting a complex merger which encompasses [w], [v], bilabial fricative [β] and bilabial approximant variants.

Dental fricatives /θ/ and /ð/ are found, but per BBerE’s patterns: word-medial and word-final /θ/ as /f/ but /θ/ word-initially; /ð/ as /d/ word-initially, /ð/ word-medially, and /v/ word-finally.

Between vowels, and before a syllabic /l/ or /m/, the alveolar plosives /t/ and /d/ are commonly realised as an alveolar tap/flap [ɾ] (e.g. in water, lady, bottom, bottle). For consistency with a not dissimilar feature of the US English model, these are reflected as /d/ in the transcriptions.

Word-final plurals are commonly shown as /s/ (e.g. shoes /ʃʉs/, wings /wiːŋs, wɪjəŋs/, misses /ˈmɪsəs/). Past tense -ed is /ɪd/, while -ing is shown as its commonly fronted form /ɪn/ (the vowel quality varies from close front to schwa, so close front qualities may be observed on audio). Following /t/, -ing is shown as syllabic /n/ (e.g. pitting /ˈpɪtn̩/).

Neither BBerE nor TBWBerE have linking or intrusive /r/. Yod-dropping after /t d n/ is commonplace in BBerE, and /h/-dropping is only noted in unstressed function words (also no /h/-insertion). BerE has no velar fricative as in British English loch (instead is simply /k/). wh- is shown as /(h)w-/, since the voiceless labial-velar fricative [ʍ] noted by Ayres (commonly shown in OED as /hw-/) was not found in BBerE by Hall. Four consonants, /m, n, ŋ, l/ are feasibly syllabic in BerE.

Word-initial clusters involving a fricative and /r/ are shown per British English and with palatalized fricative (e.g. three as /ʃri/, street as /ʃtrit/). A distinctive BBerE feature of /t, d/ deletion from the second element of syllable-final clusters is also represented here, both in monomorphemic and past-tense contexts, resulting in last as /lɑːs/, products as /ˈprɑdʌks/, old as /əʊl/, island as /ˈɑːɪlən/, deceased as /dəˈsɪːs/, bender as /ˈbænə/ and wanted as /ˈwɔənəd/ (examples based on Hall 2018).

Several features are not reflected in the pronunciations but may be noted in the audio recordings. /l/ is commonly vocalized in coda position (except at word boundaries before a vowel), but is not reflected in OED as dark /l/ is a feasible variant (though note GOAT before /l/ is /əʊ/). The glottal stop [ʔ] is the most common allophone of /t/ in word-final position, before consonants in a syllable coda, and before syllabic /n/ (though not /l/ or /m/).

Key sources for Bermudian English