Accent Reversion in Older Adults: Evidence from the Queen's Christmas Broadcasts
Jonathan Harrington and Ulrich Reubold
In recent years, there has been an increased interest in investigating the lifespan i.e., linguistic and phonetic changes during adulthood (see Bowie and Yaeger-Dror 2015; Sankoff 2018 for reviews). Some of these changes, such as a decrease and possible subsequent increase in the fundamental frequency (Linville 2001) and first formant frequency (Linville and Fisher 1985; Reubold and Harrington 2015; Xue and Hao 2003) with increasing age are a consequence of the physiological maturation to the vocal tract (Reubold et al. 2010).
But other lifespan changes can be phonetic and caused by adaptation to the community to which a speaker is exposed. This was demonstrated in a study of a bilingual English-Portuguese adult speaker in Sancier and Fowler (1997). Her English voiceless stops had shifted acoustically in the direction of their Portuguese counterparts after she had spent several months in Brazil. More recently, in a longitudinal study of the famous tennis player and late German-English bilingual Stefanie Graf (de Leeuw 2019), the phonetic characteristics of Graf’s L2-English had over several years influenced her LI-German.
Long-term change in adulthood can also be brought about by exposure to a new dialect (Chambers 1992). Kwon (2018) has shown significant non-categorical shifts in Noam Chomsky’s native Philadelphian vowel system towards the new dialect features after his relocation to Boston. In Reubold and Harrington (2018), it was shown how the vowels of the broadcaster Alistair Cooke shifted away from his standard British English accent, Received Pronunciation (RP) that he spoke at the age of 25 years and towards American English as a consequence of moving permanently to America in 1937 at the age of 29 years.
For several decades, an analytical device for the examination of diachronic development in a language was the apparent-time construct (Bailey et al. 1991; Weinreich et al. 1968), which was based on the idea that differences among generations of similar adults can be used to measure a sound change in progress because language was assumed to become more or less stable after the critical age of language acquisition. There are indeed many longitudinal studies demonstrating adult stability rather than change in the direction of (or against) the community (e.g., Bowie 2015; Brink and Lund 1975, cited in Labov 2006; Labov and Auger 1998), and stability has been reported in Sankoff’s (2018) overview to be the most common trajectory in post-adolescence. However, there are also studies showing that phonetic shifts in adulthood are triggered by sound changes sweeping through the community (Sankoff and Blondeau 2007). According to Sankoff (2018), the second most common trajectory of change within an individual adult (after stability) is a so-called lifespan change, defined in Sankoff (2005) as an individual’s modifications in the direction of ongoing community-wide change. Buchstaller et al.’s (2017) longitudinal analysis showed that some but not all of the six adult speakers from Tyneside, England, recorded in 1971 and in 2013, had followed the community changes by which the face diphthong has become more monophthongal and phonetically closer in this Tyneside variety. A longitudinal analysis by Quene (2013) showed that Queen Beatrix of the Netherlands increased her speech rate between the ages of 42 and 74 years, compatibly with the more recent tendency to produce speech at a faster rate in the Dutch language community. Based on an analysis of the annually produced Christmas broadcasts, Harrington et al. (2000a) showed how the accent of Queen Elizabeth II had shifted over a 30-year period from a conservative towards a more mainstream form of RP (Received Pronunciation). An uncommon trajectory in post-adolescent language users is a so-called retrograde change (Sankoff and Wagner 2006, 2011; Buchstaller 2016) in which adults revert to more conservative language forms that were used at an earlier time in the community. Retrograde change can apply to linguistic categories in e.g., the use of two different kinds of future tense in Quebecois French (Sankoff and Wagner 2011). Only three studies have to our knowledge demonstrated retrograde phonetic/phonological change in older adults’ speech. Shapp et al. (2014) showed that the Justice of the Supreme Court of the United States, Ruth Bader Ginsberg, increasingly used in late adulthood older (and now outdated) New York City (NYC) vernacular variants including a raised vowel in thought and vocalized rhotics. MacKenzie (2017) analysed longitudinally the frequency of intervocalic /r/-tapping in the natural historian and broadcaster David Attenborough, a speaker of RP. Although /r/-tapping has waned in this variety in the last 50 years, David Attenborough exhibited beyond the age of 74 years a retrograde change and increased /r/-tapping in later life. Reubold and Harrington (2015, 2018) showed a case of “accent reversion” in the aforementioned broadcaster Alistair Cooke who, having shifted his RP accent towards American English for certain vowels, then reverted his accent beyond the age of 68 years back in the direction of RP even though he remained domiciled in the United States during the period of accent reversion.
Shapp et al. (2014) attribute their finding of a retrograde change to Ginsberg’s positions (lawyer vs. judge) at the Supreme Court which were reflected in the different ways that she accommodated to non-NYC English speakers. MacKenzie’s (2017) explanation for Attenborough’s retrograde retention of intervocalic flaps is that these have been lex- icalised within frequently occurring collocations such as “for a” and “their own”. In Reubold and Harrington (2018), the accent reversion in Alistair Cooke’s late adulthood was found to be led by high-frequency words. Their tentative explanation for this finding is that more recently learned exemplars of words, as well as those of low frequency (Bybee 2006), are less entrenched in memory; and, in very late adulthood, it is the less entrenched exemplars that are deleted first. However, this conclusion is highly speculative, being based on evidence from one person (Alistair Cooke) and for one vowel change (in which an RP-like bath vowel with a back quality close to cardinal vowel (CV) 5 shifted due to exposure to American English towards the front of the vowel space but then back again towards CV 5 in late adulthood).
In the present study, we seek to shed more light on the issue of retrograde change and its possible interaction with lexical frequency through a reanalysis of the Christmas broadcasts that have been delivered annually by Queen Elizabeth II. The focus is on three vowels for which we had earlier found evidence of a phonetic change in the direction of a more modern, mainstream RP (Harrington et al. 2000b; Harrington 2006). These include the /и/ as exemplified by the lexical set goose, the Ixl of trap, and the M of happY which were found to be phonetically more front, lowered, and more peripheral in the later decades (in the 1980s for /и, a/; up to 2002 for M) compared with recordings obtained from the 1950s. The main issue to be considered in the present study is whether in later years up to 2017 these vowels of the Queen’s Christmas broadcasts have remained stable, or whether there is evidence of a retrograde change towards those qualities from the 1950s combined with an interaction with lexical frequency.