19th century sound change in Kent: TRAP


The distribution of the [æ]-like TRAP vowel by the eight informants: this is either the earlier timbre close to DRESS (O), or it is the new open timbre (N). Two informants still had the earlier pronunciation.

The map shows that the new open [æ]-like pronunciation dominated most of Kent, with just two still acquiring the older closer [ae]-like pronunciation near DRESS. Six informants had acquired the new pronunciation by 1865-1895, the Table below suggesting that perhaps 25% of the county still retained the old pronunciation. This means that the origin of this change is probably similar to that of STRUT, starting perhaps around 1810-1820.

This post continues from the introductory Nineteenth century sound change in Kent, where the informants are introduced and the background presented.


The numbers of informants with new pronunciations (N) and old pronunciations (O), for each of the ten sound changes, ordered from the left by the total number of O acquired, assuming a progression from the earliest to the most recent. The grand total for each sound change is 8, the number of informants. The new open [æ]-like TRAP was acquired by six out of the eight informants, and the earlier close [æ]-like pronunciation by two.

Alexander Ellis (1889, On Early English Pronunciation, Vol. 5, Teubner) recorded the surviving dialect pronunciation of TRAP in Kent and East Sussex as [æ], written in his palaeotype alphabet. He described this as low front wide, following Bell, adding it “approaches close to E”. This puts it at IPA lowered [ɛ], where the IPA also has æ. Eustace drew the same conclusion (1969, The meaning of the Palaeotype in A. J. Ellis’s On Early English Pronunciation 1869-89, Transactions of the Philological Society 68(1):31-79). Wyld explained that TRAP landed near DRESS as the outcome of the original TRAP-BATH split (1936, A History of Modern Colloquial English, Oxford, Blackwell), and in London East End Cockney it even moved to [ɛ]. In either case, the consequence was compression of DRESS and KIT towards FLEECE.

Nevertheless, Ellis classified it as low front, despite the proximity to DRESS. I would also prefer to see an [æ]-like zone filling low front (i.e. high F1 and high F2), with the older closer [ae] at the edge near DRESS, and the new open [æ] in its centre. Along with the new open [æ], KIT and DRESS spread out more. The new open [æ] is sometimes transcribed as cardinal [a] in the literature. Daniel Jones exemplified his cardinal [a] with “Parisian French a in la”, which is definitely not the same timbre as the new open [æ]. When Daniel Jones modified the Bell model for his cardinal vowels, he adapted it to Paul Passy’s French vowels by reducing Bell’s three low vowels to Passy’s two. French [a] became the frontmost vowel, leaving open [æ] at more front than front, with higher F2 (Wood, forthcoming, Cardinal 4 is open central and not open front).

F1/F2 diagrams for TRAP in 19th century Kent
showing older (O) and new (N) pronunciations

The timbre of the earlier TRAP in 19th century Kent is closer to [ɛ]-like DRESS. The formant diagrams above show that six of the informants (abcefh) had adopted the new shifted TRAP timbre with F1 higher than 600Hz. The other two (dg) had retained the earlier pronunciation with F1 lower than 600Hz. Additionally, informants (dg) have KIT and DRESS compressed towards FLEECE, with F1 300-500Hz. Informants (abcefh) had KIT and DRESS spread out with F1 400-600Hz.


Sound examples

The earlier close TRAP near DRESS

The new open TRAP

Examples from contemporary RP

For comparison, here are two RP examples, contemporaries of the Kentish informants. One is a political broadcast recorded by the BBC in 1931 by politician Stanley Baldwin (1867-1947), and the other is a Linguaphone recording (British Library) made in 1929 by Professor Daniel Jones (1881-1967). Henry Sweet (1877, Handbook of Phonetics, Oxford, Clarendon), Daniel Jones (1932, Outline of English Phonetics, and later editions to the 1960s) and Gimson (1962, Introduction to the Pronunciation of English) gave TRAP close to dress as the corresponding RP pronunciation.  Gimson reported the open variant as regional. Wells reports the open TRAP as a recent innovation in RP (1982, Accents of English Vol.2).


F1/F2 diagram, showing the new open TRAP in RP

Surprisingly, the two RP examples have the new open TRAP, with F1 higher than 600Hz, and KIT and DRESS consequently spread out (F1 400-600Hz). This is a century before this sound change was first detected and one wonders how it came to be missed, since it was readily recognized in regional speech.

©Sidney Wood and SWPhonetics, 1994-2016


This entry was posted in Accents, Articulation, Dialects, English, Kent, Pronunciation, RP, Vowels and tagged , , , , , , , , . Bookmark the permalink.