1. The Problem
|Figure 1. The earlier closer TRAP, with DRESS and KIT compressed towards FLEECE. RP informant B born around 1900.|
|Figure 2. The new open TRAP; F1 completely higher than 600Hz; DRESS and KIT are no longer compressed towards Fleece. RP informant Daniel Jones, born in 1891.|
|Figure 3. The stylized isogloss between Northern British English (NBE) and Southern British English (SBE). Four regiolects of SBE: East Anglian, London, Home Counties (triangle) and Western.|
A recently observed sound change (Wells 1982) in southeastern England opened1 TRAP2 pronunciation both in non-regional Received Pronunciation3 (RP) and in regional accents. The timbre of the earlier closer TRAP was IPA [æ]4 (Fig. 1), while new open TRAP is open [æ] (Fig. 2). The criteria for distinguishing new open TRAP, F1 higher than about 600Hz and no compression of DRESS and KIT towards FLEECE, will be explained later. The regional Kentish TRAP sound change had commenced by the 1860s (Wood, 2017). It was also spreading through the home counties (the region surrounding London and bounded by East Sussex, Hampshire and Northamptonshire, Fig.3). The earliest RP speaker with new open TRAP reported by Fabricius (2007) was born in 1926. Did Fabricius catch it starting in RP during the 1920s, or was it already in progress? Figure 2 suggests it was (Daniel Jones born in 1881). Early examples like this are considered controversial, born long before anyone showed any awareness of this sound change in RP. Jones never mentioned it himself, always describing only the earlier closer RP TRAP timbre (1908, 1918, all editions to the 1960s). An additional recording of Jones has recently been analysed by Przedlacka & Ashby (2019). Their vowel diagram (their Fig. 1) also shows Jones’ open TRAP at high F1. Controversial as they may be, it was examples like this, together with regional examples from SE England, that first prompted doubts about the dating of this sound change.
The status of new open TRAP in earlier RP was investigated by analysing the spectra of TRAP instances taken from recordings of RP speakers in two sequences preceding Fabricius’ confirmation: (i) an RP-speaking group born in 1850-1899 (the controversial period), and (ii) a second RP group born in 1900-1930 (the non-controversial period up to Fabricius’ dating). For comparison, published formant data from regional Kentish SBE speakers born in 1860-1895 (Wood, 2017), and recordings of six speakers of 20th century Home Counties SBE (HCSBE) born in 1900-1960 were also included.
Continued at 2. Accents of SE England
1 For the present study, the terms tongue height and backness etc. follow the usage of the past 150 years (A. M. Bell 1867:15-16, 71, Sweet 1877, D. Jones 1932, IPA 1999), but advisedly because the Bell vowel model has never been validated, but was seriously compromised on numerous occasions (Wood, 1982).
2 Expressions like TRAP are keywords for what J. C. Wells (1982:§2.2) calls lexical sets representing vowels that participated in various sound changes in English, with different outcomes in different dialects. They are more useful than phoneme notation when pronunciations are changing, or where accents differ.
3 The expression Received Pronunciation (RP) is preferred here for the sake of continuity with earlier literature although other rival synonyms have come into use for this accent, especially General British English (GBE) (Lewis 1972) and Standard Southern British English (SSBE) (IPA 1999). SSBE is especially unsuitable as it is also increasingly being used as a nickname for regional HCSBE, the “new standard” (like Lindsey 2019:4). The phonology of non-regional RP nevertheless belongs typologically to SBE.
4 Close [æ] refers to IPA [æ], the timbre of old close TRAP. The timbre of open TRAP was never given an IPA identity and is referred to here as open [æ].