Revision 1: 13 June 2018
(i) revised judgement on RP09’s MOUTH vowel
(ii) added a second example, RP10, whose MOUTH vowel was similar to RP09’s.
These two examples are taken from the sound track of a BBC discussion. One, referred to as RP09, is a well known broadcaster. The other, referred to as RP10, is a professor. Private data is withheld for both for the sake of privacy.
Listen first to these brief extracts from each.
“… now that is, that is a move forward, there are, uh, international organizations in the United Nations, who are also, there is an environmental, program …”
“… part of our problem, is poverty, human poverty, and part of our problem is, human ignorance, and both of those conditions were manifest, in many parts, of the w.., now so-called, western world, just after the second world war, and many people in this room, and others, will remember, that the Thames was filthy, the cities were dirty …”
The recordings were analysed by linear prediction, using formant tracking in Praat. The formant frequencies of vowels were collected from fully focused syllables, to avoid spectral variation due to vowel reduction. Vowels were sampled at the moment where vowel articulation was least affected by surrounding consonant articulations, selected by comparing VC and CV formant transitions, in order to minimize spectral variation due to coarticulation.
The RP accent is most likely to be confused with the standard sociolects of regional Home Counties or London SBE, which are similar to RP in many respects and have existed alongside RP for at least 100 years (20th century regional HCSBE had evolved from 19th century sound changes in Kent and Essex that spread through the home counties and are still spreading westwards). This similarity wouldn’t have been a problem in the past owing to the former stigma attached to regional SBE and the expectation that its speakers would adopt RP to improve careers. But there is no stigma today and speakers of regional SBE are no longer adopting RP, hence the increasing risk of confusion. The RP accent is defined by it’s phonology, as described by Jones (1932-68) and updated by Gimson (1962) and Wells (1982, Vol 2). Several sound changes have occurred in RP, and 20th century RP speakers should have acquired the most recent – STRUT from [ʌ] to [a], TRAP from close to open [æ], and GOAT from [ou] to [əʊ]. A small set of vowels is a good starting point for distinguishing between RP and regional HCSB – LOT (RP [ɒ] and HCSBE [ɔ]), THOUGHT (RP [ɔː] and HCSBE [oː]), MOUTH (RP [aʊ] and HCSBE [æɒ]), and GOAT (RP [əʊ] and HCSBE [aʊ]). Note that RP MOUTH and HCSBE GOAT sound very similar. Increased GOOSE fronting is also a strong tendency in 20th century RP (this was actively avoided in earlier RP).
Figure 1. F1/F2 diagram for monophthong vowels by RP09
|Figure 2. F1/F2 diagram for monophthong vowels by RP10|
These diagrams illustrate monophthong vowels by RP09 and RP10. Their formant frequencies are unusually high, indicating shorter than average vocal tracts for both. F1 for RP09’s open vowels TRAP, STRUT and BATH is 650-950Hz, F2 for FLEECE reaches beyond 2500Hz. F1 for RP10’s open vowels extends beyond 800Hz, F2 for FLEECE reaches 2400Hz. Unrounded open is typically defined as F1>600Hz, while this limit is around 650Hz for RP09 and nearer 700Hz for RP10.
Both have all the RP sound changes. TRAP and STRUT are open. GOOSE was fronted (F2 extending to 1300-1500Hz by RP09 and 1300-1700Hz by RP10). RP09 also had some instances of GOOSE as low as 800-1000Hz (where RP used to have all its GOOSE renderings). They both had preglottalized unvoiced stops in some locations, that would have been stigmatized as regionalisms in the past, but which have nevertheless occurred in RP since at least RP09’s and RP10’s generation.
Figure 3. The start and end zones for the diphthong GOAT (RP [əʊ]) by RP09.
|Figure 4. The start and end zones for the diphthong GOAT (RP [əʊ]) by RP10.|
These diagrams above illustrate the GOAT pronunciations, demonstrating RP [əʊ] rather than HCSBE or London SBE [aʊ], where the start F1 would have been higher and around 700-800Hz. There was one outlying instance starting at [a] by RP09 that might have been a sporadic regional pronunciation, but it’s disregarded here as spurious and unexplained in view of the small sample.
Listen to these instances of the RP [əʊ]-like GOAT vowel by RP09 and RP10:
RP09: “so, so, growth, those, don’t”
RP10: “approach, both, go, no, no, only ”
These diagrams above illustrate the MOUTH pronunciations, demonstrating an unexpected [aɒ]-like timbre, rather than the typical RP [aʊ] (for comparison see the ending of the GOAT diphthong in Figures 4-5, and the LOT zones in Figures 1-2). These pronunciations can hardly be explained as reductions. They are all taken from fully focussed syllables and consistently conform to the same regression. The start timbres, however, are typically RP [a] (although RP09 did produce two instances starting at regional HCSBE or London [æ], ignored for now as his sample is so small)
Listen to these instances of the MOUTH vowel by RP09 and RP10:
RP09: “how, allow, mountain, doubt”
RP10: “now, out, about, now”
This the first time I have found a consistent [aɒ]-like timbre for MOUTH in otherwise typical RP accents, and I haven’t seen it reported elsewhere. It might be a regionalism, but that would be purely guessing since nothing is known of these informants’ linguistic history. It could equally be the first sign of a new sound change in RP, or simply idiosyncratic.