These notes are largely based on Lodge (Reference Lodge2009), and the IPA transcriptions on Wells (Reference Wells2000).

Michael Bulley has singled out for discussion one variant in the articulation of the /r/ phonemeFootnote ¹ in British English which he characterises as ‘w-for-r’, ‘close to /w/, but distinct from it’ (Bulley, Reference Bulley2014: 46). The focus is on /r/ in syllable-initial positionFootnote ² , including /r/ in initial consonant clusters. He describes this particular, not to say singular, variant as ‘/w/-like’, which begs a number of questions.

As a set of such diverse variants, /r/ is an odd enough phoneme in British English. The variants are determined principally by contrasting postures of the tongue in close proximity to the roof of the mouth:

• • generally, an alveolar trill [r], with the tip of the tongue raised to the back of the teeth ridge, though more usually an approximant articulated simlarly but without trill [ɹ];

• • or a retroflexive articulation with the tongue tip flexed backwards [ɻ], typically in rhotic accents giving a preceding vowel ‘r-colouring’ in overlapping articulations: ‘firm’ [fɝ:m] as opposed to non-rhotic [fɜ:m]; syllable articulations are sequential rather than segmented;

• • alternatively a tap or flap [ɾ] - with an added complication that this is also an alternative to the aspirated realisation of /t/ between two vowels: ‘bitter’ /'bɪt^hə/, or ['bɪɾə] confusable with ‘bidder’ /'bɪdə/;

• • locally, as a minority form in North-East England and parts of Scotland, a uvular articulation with the back of the tongue raised to the velum or soft palate, again either with trill [ʀ] or without [ʁ], similar to /r/ in French and German.

What certainly is odd is the variant seemingly identified by Bulley, an altogether different labiodental articulation, a close approximation of the lower lip to the upper teeth, [ʋ]. Phonetically this is more /v/-like than /w/-like, though not to be confused with the phonologically unrelated /v/, a labiodental fricative with actual closure and audible friction, and lexically contrastive: ‘rigour’ ['ʋɪɡə] – ‘vigour’ /'vɪɡə/, ‘rice’ /raɪs/ – ‘vice’ /vaɪs/ .

So why ‘/w/-like’? This is to do not with the articulation, the phonetics, but with the sound systems of the language, the phonology; */gv/  is not a possible initial cluster in English, while /gw/ obviously is, akin to the voiceless /kw/ spelt as <qu-> (quick, quiet…), if only for a handful of ‘Celtic’ proper names (Gwendolen, Guinivere, Gwynneth, Gwent…) Hence, to take Bulley's example, for ‘grown up’ a child-like articulation as ['ɡʋəʊnʌp] would hardly sound like an impossible *['ɡvəʊnʌp] but more like ‘gwown up’, hence ‘w-for-r’.

But that is the other oddity of [ʋ]; it is a personal, not a regional, local or contextual variant, a developmental form of /r/ in child language, and in an individual adult's speech a feature sometimes described as a speech defect. It attracts attention in the speech of such public personalities as British TV presenters naturalist Chris [kʋɪs] Packham and historian [hɪ'stɔ:ʋiən] Lucy Worsley, and may well have been a variant particular to RP, with ‘posh’ and ‘privileged’ connotations. Lodge (Reference Lodge2009, p. 180) notes:

A post-script, prompted by the graphic gag in Bulley's title, ‘W(h)ither… ?’ In Scottish accents the initial articulation with the corresponding spelling <wh-> is voiceless. Is this just a variant [ʍ] of the phoneme /w/, voiced in other accents, or indeed a separate phoneme /hw/ (minimal pairs: watt-what, weather-whether, wither-whither, wear-where, wail-whale…)? Discuss.