We are neither automatons nor untrammelled agents when it comes to sociolinguistic style. The sung voices of popular music exemplify this ever-present tension between convention and innovation, between belonging and branching out.

Pop Song English (PSE) is a supralocal norm that, despite being based on the phonology of American Englishes, is performed in Anglophone music around the world. It has been conventionalised and focused over the last six or so decades, and despite its likely origins in intentional stylisation (as evidenced by instances of overshoot), it has become the default style for singing pop songs. While cases of ‘own-accent’ singing attract attention, their salience is in part due to the unmarkedness of PSE.

In a corpus of 154 artists, structured according to country of origin (NZ and USA) and musical genre (pop and hip hop), auditory analysis of BATH vowels and nonprevocalic /r/, and acoustic analysis of LOT vowels reveal greater homogeneity in pop than hip hop, and show how closely NZ pop singers align with their US counterparts. Against this backdrop of conformity, artists who wish to ‘sing the way they speak’ may do so with conscious intent, effectively stylising their own spoken voice in the context of song.

The formation of new musical genres involves a range of stylistic innovations that sometimes include the development of singing accents that break away from PSE. In this way, musical genre (rather than geographic origin of a singer) becomes the primary predictor of singing accent.

Moving from the sociolinguistic to the psycholinguistic, I argue that PSE is not restricted to performers per se. The results of two perception experiments suggest that, in the New Zealand context, the music-listening public undergo a perceptual style shift when expecting to hear a sung voice. They categorise phonemes according to a more American-like phonological frame of reference, and process words more quickly in a US than a NZ voice in a musical context. Singing and speech constitute sufficiently distinct domains to support a kind of bidialectalism for listeners and singers alike.

In 2014, then Education Secretary Michael Gove led education reforms that included the reintroduction of explicit knowledge of English grammar in the National Curriculum for the first time in half a century. Over the last decade, the Englicious project, based in the Survey of English Usage at UCL, has produced resources aimed at helping pupils, teachers, and parents with these new requirements. These include the free website englicious.org as well as classroom materials and training sessions for school staff.

The inclusion of grammar, especially the testing at primary level and some new terminology, has proven to be persistently controversial, epitomised in the press and social media bugbear, the ‘fronted adverbial’. This talk will cover the recent history of grammar in the curriculum, and discuss how the Englicious project has helped teachers gain a better understanding of grammar pedagogy.