1In his pioneering work on popular music productions of the 1960s and 1970s, Trudgill reported a tendency for British singers to “employ different accents when singing from when they are speaking” (141). He identified a number of phonetic variables in the sung productions of such artists as The Beatles and The Rolling Stones, which commonly occur in American accents and sound stereotypically American to the ears of British speakers. This led him to conclude that this modification on the part of British artists is a form of Americanization of singing-voice pronunciation. Rather than a change of accent as such, this phenomenon can be understood as an example of style-shift, as defined by Labov: speakers (in this case, singers) vary their pronunciation, adopting styles they deem relevant to specific situations. This shift is interpreted by Trudgill as an example of an “act of identity,” referring to Le Page’s theory of linguistic behavior. As "linguistic chameleons," (Le Page 18) individuals can modify their linguistic systems and behavior to match a group they wish to identify with at a given time – where identification is understood as a lifelong process of creation “according to various social constraints (historical, institutional, economic, etc.), social interactions, encounter, and wishes that may happen to be very subjective and unique” (Tabouret Keller 316). In other words, because of the inherent link between identity and language, individuals create and project an identity through various acts of language (Le Page et al.). Applied to British artists, adopting an Americanized pronunciation in a communicative situation of performance allows them to identify themselves with an American model group. According to Trudgill, this is motivated by the cultural domination of the United States in the field of popular music at the time. Most, if not all, popular music in the Western world in the 20th century directly stemmed from (Afro-)American music (Trudgill, Morrissey, Westphal and Jansen, Frith). As such, it was appropriate for British singers to sound American when singing, since they were taking part in a chiefly American activity. This was true for most of the 1960s and 1970s productions studied in Trudgill’s work and had been occurring since as early as the 1920s. Despite this, Trudgill argued that the advent of the First British Invasion in the United States weakened the strength of American domination in popular music. This cultural phenomenon of the mid-1960s saw a rise in popularity of British artists in the United States, in part thanks to the overwhelming success of The Beatles (Perone). British pop music consequently acquired validity with some British artists switching back to linguistic variants more consistent with their geographical origins.
2Nevertheless, there remains an undeniable use of some type of American-influenced singing style on the part of British artists in subsequent decades, which some studies on the sociophonetics of the singing voice have examined. Simpson followed up Trudgill’s study looking at later examples of pop and rock singing, adding considerations linked to stylistics and register. Morrissey further explored this tendency in selected examples from British pop and rock. Konert-Panek focused on specific pronunciation traits of artists such as Adele and Amy Winehouse, highlighting a similar trend. There have also been a number of similar studies on artists from other backgrounds, including Gibson on New Zealand popular music, Duncan on Australian country singer Keith Urban, or Krause and Smith on the Glaswegian indie music scene.
3The approach to style-shift in song from the perspective of Acts of Identity as suggested by Trudgill has since been updated, notably by the introduction of the theoretical construct of Audience Design (Bell, Bell “Back in style”), which sees style as a speaker’s response to their audience and context. American-influenced singing, then, could be seen as a responsive style (Gibson and Bell, Gibson “Pop Song English”) because of “its predictability in, and appropriateness to, the pop song context,” (5) due to the influence and domination of the United States in the development of recorded popular music. The implication is that though in the past there may have been a more conscious and overt American influence on the pronunciation of British singers, this American-influenced style has grown to become enregistered, or socially recognized as indexical (Agha) of pop singing, divesting it of some, if not most, of its geographical connotations. In other words, earlier American influence has evolved into what is considered an appropriate pronunciation style for pop singing, now a largely unconscious standard adopted by British artists and others. Beal studying the Arctic Monkeys’ use of their native Sheffield accent in song argues that it is in reality the performance of one’s own accent that requires conscious attention rather than an American-influenced style, which has become the default singing pronunciation. Further perception evidence supports this stance, with British listeners identifying an Americanized accent as the expected style in pop songs (Jansen). This variety of English in its own right is referred to as Pop Song English (PSE), defined as a supralocal norm for singing in popular music (Gibson “Pop Song English”).
4The timeline of the evolution from an explicit attempt to ‘sound American’ towards the enregisterment of this pronunciation as Pop Song English is imprecise by nature. Most authors agree that 20th century popular music was directly influenced by an ‘American’ model, while the shift towards a pop-singing standard was finalized in the 21st century. 1980s pop music, which this article focuses on, thus represents somewhat of a grey area. On a cultural level, this decade was a major turning point in the production, distribution, and reception of popular music (Frith, “Music for pleasure,” Negus) and saw the creation of a ‘new mainstream’ with the return of the Top 40 radio format (based on a precisely curated and heavily repeated rotation of the most popular hits of the moment), the emergence of MTV revolutionizing the whole industry (Straw), and the rise in popularity of electronic dance music epitomized by the synthesizer (Goodwin). All this led to the Second British Invasion in the early-to-mid 1980s (Puterbaugh), with British acts such as The Police, Duran Duran, or The Human League becoming unavoidable hit makers in the US.
5In addition to purely dialectal considerations, there have also been a few attempts at looking at this type of style-shift through a more phonetic, phonological, and articulatory lens. Morrissey suggests this type of pronunciation change may occur in singing simply because some phonemes “lend themselves better to singing than others” (209), an idea which he refers to as ‘singability.’ Namely, he argues that there is a tendency for greater sonority in song and that several of the American features favored in singing by British artists have a higher sonority than their British counterpart and are therefore more ‘singable.’ Gibson also underlines the importance of singing technique, and the fact that greater jaw opening leads to more open vowels and therefore to higher F1 values. However, this is an aspect which has rarely, if not at all, been explored in depth in most sociophonetic research on singing in popular music.
6In his study, Trudgill listed several pronunciation features adopted by British singers which fit a broad and non-specific American accent, later conceptualized as the USA-5 model (Simpson), summarized below:
1) The pronunciation of intervocalic /t/ in words like better or city as [d] rather than [t] or [ʔ] commonly used by British speakers.
2) The long open vowel [ɑː] in words like dance or last of Southern British English tends to be realized as a shorter and broader [a].
3) Non-prevocalic /r/ is pronounced, even when the singer’s regular speech is non-rhotic.
4) The diphthong [aɪ] in words like life or like is realized as [a⋅].
5) A process similar to LOT-Unrounding occurs where words like body and top are pronounced with [ɑ] rather than a rounded [ɒ].
- 1 In this paper, what we call American English corresponds to Wells’ General American category. Br (...)
7Among these phonetic variables, the spoken pronunciation of 1) intervocalic /t/ marks a strong distinction between British and American English1 (De Jong, Wells), and has been described as “one of the most striking characteristics of American pronunciation to the ears of a non-American” (248).
8Before further detailing the pronunciation of intervocalic /t/, a terminological caveat is in order. Articles on this typically American phenomenon tend to call it one of three ways: T Voicing (Wells), Flap (Braver, De Jong), or Tap (Alderton, Shockey). The latter two refer to specific phonetic realizations that remain to be more precisely delineated (De Jong, “Stress-related variation”). The former, T Voicing, is a more phonologically oriented label, though voicing is not a necessary requirement for the production or perception of a flap/tap since “the most salient acoustic feature of the tap is not (…) its voicing—which need not even be present—but rather its shortness” (Trask as cited in Iverson and Ahn). Voicing itself can be seen as an optional secondary step after a flapped articulation in the process of full neutralization of the /t/-/d/ contrast (Wells). Despite their subtle differences, in most of the literature these terms are used interchangeably to describe a similar phenomenon. This terminological variation relates to the interface between phonetics and phonology, and the difficulty of coherently labeling the same speech sounds from one perspective or the other. For the sake of clarity, this phenomenon will henceforth be referred to as flapping as it is the label used most frequently in the literature (De Jong) and flaps are the preferred realization in American English speakers’ productions (Derrick and Gick), though taps occur as well.
9It must be noted that flapping can be inconsistent in its realization even for American speakers, and that the entirety of the conditions ruling its production are not fully understood (Shockey). Nevertheless, the generally agreed-upon rule has not changed since one of its first formulations in the 1930s: “[flapping] cannot occur unless it is preceded by a vowel or a sonorant (n, l, r), and is followed by an unstressed syllable-forming element (vowel, l, r, but not homorganic n)” (Haugen 631). For the sake of simplicity, we will keep referring to /t/ in such places as ‘intervocalic,’ though as the definition suggests, this also includes sonorants. Furthermore, the focus will be limited to the pronunciation of intervocalic /t/ in word-internal environments (where flapping is possible) only. Flapping can also happen across word boundaries as a linking process (a lot of; get on) but was excluded from the study as it can be heard often enough in British English – especially in Cockney (Wells) – and is therefore not as distinctively an American trait as its word-internal counterpart. Nowadays, word-internal intervocalic /t/ flapping can be heard among younger speakers of Standard Southern British English, though to a lesser degree than other allophones of /t/ (Alderton). Nevertheless, this work assumes that this pronunciation trait in speech was markedly American at the time of Trudgill’s analysis (Wells) and, as such, during the time of production of the artists under study in this article. As a trait that is representative of a generic American accent (as opposed to a British one), it may therefore have been a preferred alternative in early pop song pronunciation. Over time, however, it has come to be considered in singing not just as ‘American’ but also as indexical of PSE (Beal).
10The aim of this paper is threefold:
11Given the context of the 1980s, and Trudgill’s argument that the First British Invasion legitimized the use of a British accent in pop singing to a certain extent, did the Second British Invasion have a similar effect? The first aim is therefore to carry out a large, quantitative analysis of popular music productions in the 1980s to get a sense of the distribution of /t/ flapping as a pronunciation indicator of an American-influenced singing style or PSE.
12Whether 1980s pop artists were still under the active influence of the American model or already leaning towards the use of PSE as a singing pronunciation norm, sociocultural factors pertaining to the domination of the American music industry and the authority of American music genres play a certain, more or less direct, role in the singing pronunciation of British artists. The second aim of this paper is therefore to explore such factors and use artists’ own perspectives to complement these points.
13Lastly, if phonetic factors have previously been suggested as a further aspect to take into consideration, they have often been glossed over in the literature. The third aim of this paper is thus to suggest an alternative and complementary articulatory take on the question, based on the specificity of the singing voice whose inherent constraints cannot be overlooked in the development of Pop Song English as a pronunciation norm. In other words, PSE’s early influences might have been clearly American, but there might be an additional, articulatory argument as to why it evolved into a standard, specifically regarding the pronunciation of intervocalic /t/.
- 2 Some albums may have been released in 1978 and 1979 but topped the charts in 1980. They were the (...)
- 3 See the Appendix for the full list of artists, albums, and songs that were included in the analy (...)
14Using Trudgill’s study as a baseline, we wanted the database to be as representative a sample of the 1980s as possible and large enough to allow robust quantitative analyses. While Trudgill selected a few albums by a few acts, other scholars studying the same phenomenon relied on even less source material, sometimes only analyzing a single chorus from a specific artist. Though far from being the sole indicator of success, we relied on sales numbers as a starting point. For each year of the 1980s decade (including 1990)2, we selected when possible the first ten best-selling albums in the United Kingdom by male English artists. Inclusion criteria were the following: artists had to be male to avoid any physiological or gender-related variation; they had to be English specifically from England – not British, or not English having been born and raised elsewhere in the world – and albums that were too marked outside the general umbrella of pop, such as reggae, as well as live albums, and compilations that might have included songs from before the time period under study were all discarded. This left a total of 86 albums by 54 different artists, including the likes of Depeche Mode, Elton John, or Wham!. From this collection, 350 songs were randomly selected to have a more manageable sample for analysis. They were then manually verified: when a song turned out to be an instrumental track, had very few lyrics, lasted less than one minute, or was a cover of an original track by another artist, it was replaced with another randomly selected song. The final sample for analysis thus consists of 350 songs from 85 different albums recorded by 49 artists over a ten-year period3.
15Prior to analyzing the pronunciation of word internal intervocalic /t/ in the final database, the 350 randomly selected songs were run through the music source separation tool Spleeter (Hennequin et al.). Spleeter takes regular mixed-track recordings and separates the tracks, effectively isolating the singer’s voice from the rest of the instruments. It is fast, easy to use, and has yielded the best results among similar tools for source separation, especially in terms of signal to distortion ratio (SDR) (Louis Ansal and Ancy). The use of this software in the study was mostly exploratory, to assess its effectiveness in isolating vocals and if it could be beneficial to the auditory analysis. The results were ambivalent: Spleeter worked very well for some songs, but for others the audio quality was notably reduced, and it tended to induce some artifacts, such as a distortion of the original signal, or instrumental lines that were not fully dissociated from the vocals. This might have had to do with the initial quality of the original tracks and the added effects such as compression or reverb that could only be worked through approximately by the software. Nevertheless, Spleeter was overall helpful in the auditory analysis, allowing the authors to hear the few ambiguous tokens of the database more clearly.
16Once all the vocal tracks from the 350 songs were isolated, tokens had to be identified. The lyrics for the randomly sampled songs were retrieved online, through a Google search. When searching for lyrics on Google, the first search result automatically uses the database of the lyric licensing company LyricFind. All words containing a word internal /t/ were automatically located in the lyrics, then manually inspected to see if they could potentially be flapped, according to the flapping rule stated in Section 1.2. If any doubts remained, the pronunciation was checked in the Longman Pronunciation Dictionary (Wells), as an indication. The stressed_unstressed or unstressed_unstressed sequences between which the target /t/ token occurred were segmented by hand in Praat (Boersma and Weenink). The authors then conducted an auditory analysis to determine if the token was in fact a stop or a flap. Glottal stops occurred extremely rarely in the data but those that did were categorized as stops since they are just as distinctly British as their plosive alternatives. The analysis was done with the deliberate goal to reach perfect agreement, based on the assumption that, at least perceptively, the two /t/ allophones would be quite clearly distinct.
17Concurrently to the investigation into the realization of intervocalic /t/, another informal auditory analysis was carried out on interviews of the artists under study, to provide a basis for comparison between their speaking and singing voice. For each artist in the corpus, three interviews (when available) spanning the whole of the decade were retrieved from YouTube. They were selected based on release year and audio quality and carefully listened to see if any spontaneous flapping occurred in the singers’ speech. Additionally, three contemporary (2010 and later) interviews were also retrieved from YouTube for each artist. Transcripts from YouTube’s automatic closed captioning option were collected for the six interviews of each artist. This allowed to search for certain keywords such as America, England, Accent, Pronunciation, MTV, etc. in AntConc (Anthony) to gain some perspective on the artists’ own attitudes towards their work in relation to some kind of American influence (or lack thereof). It is worth noting that this data is inherently observational and anecdotal, as opposed to a systematic sociolinguistic interview or questionnaire – but this was hardly conceivable due to the celebrity status of these artists. As such, it is rare that they explicitly talk about their accents, but this method (albeit not as systematic as one would wish for) nevertheless might provide some interesting additional insights.
- 4 For other features and a more artist-focused perspective, see Caillol and Ferragne for a compara (...)
18Little to no word-internal flapping occurred in the spoken interviews of the 49 different artists included in the final database. In sung productions, out of 937 words in which intervocalic /t/ could be flapped, 571 were flapped, or a total of 61%. A binomial test confirmed that there was a significant bias in favor of the flapped allophone (p < .001) in the overall database. However, when looking more closely at each artist, the proportion of flapping is highly variable. For an average of 20 tokens per artist, it ranges all the way from 0% flapped tokens for Heaven 17 to 100% for Electric Light Orchestra or Peter Gabriel. Flapping rates for each individual artist are shown in Figure 1. Despite this variation, a majority of artists (30 out of 49) flap 50% or more of their tokens, and almost all present this pronunciation trait in their singing productions, though it occurs very rarely or not at all in their speech. The study of a single feature is not enough to draw broad conclusions as to the adoption or not of an American-influenced pop-singing pronunciation norm, what is more in the productions of a large number of different artists4. Nevertheless, considering flapped /t/ only, one might conclude that it is (to a certain extent and varying according to each artist) the favored realization of intervocalic /t/ in 1980s popular music.
Figure 1. Percentage of flapped intervocalic /t/ for each artist in the database.
19Subsection 3.2. will provide an overview of typical sociocultural arguments looking at the extent of American influence and/or a PSE norm on music productions in the 1980s. The aim of subsection 3.3. is to go beyond these considerations, looking at the specificity of the singing voice. As such, we will delineate the outlines of additional articulatory factors that must be taken into account when trying to understand why PSE has become standard in popular music.
20Even if the exact moment of transition from an overt attempt to take on pronunciation features indexed as American to the enregisterment of these features as the default accent in pop singing, arguments of a broader sociocultural nature should be mentioned as part of the explanation for the style-shift English singers undergo from their speaking to singing voices. A study of this nature must consider the wider socioeconomic, cultural, and financial contexts, which undeniably act as catalysts in particular singing and pronunciation styles (Simpson, Westphal and Jansen). Trudgill suggested that the popularity of an American-influenced accent in singing might come from what he calls the American domination of the music industry. For many artists at the time, the US represented an ideal of success and accomplishment: breaking into the American market meant having made it as a musician. As Andy McCluskey from Orchestral Manoeuvers in the Dark (100% flapping) put it: “We were really trying to break America […] we had American management, we were spending a lot of time in America, we’d finally got onto a label that wanted to release our records, we'd started to have hits and we'd been immensely successful with the song If You Leave.”5 This is true if only on a numerical basis in terms of sales figures. The British Phonographic Industry6 platinum certification level in terms of sales is 300 000, whereas the Recording Industry Association of America7 places the platinum level bar at one million sales. This is relative to vastly different population sizes but does give an idea of the sheer extent of the US market and the commercial advantages that may come with it (Negus). Nevertheless, no significant correlation was found between artists’ flapping rates and their average record sales in the US and the UK. Sales numbers were retrieved when available from the official websites of the Recording Industry Association of America and the British Phonographic Industry. When albums were not certified because of low sales figures, the equivalent to the number of sales for the lowest certification was attributed nonetheless. Spearman’s rank correlation coefficients were computed, where for N = 49, ρ = 0.25 (p = .904) for the relationship between flapping rate per artist and UK sales and ρ = -0.02 (p = .085) for the relationship between flapping rates per artist and US sales. Coupled with the substantial rise of popularity of British artists in the United States with the Second British Invasion, this may have provided motivation enough to both up-and-coming and already established British acts to appeal to the US market, and perhaps even try to sound American (consciously or not) as a strategy.
21The inclination towards hypercorrection has also been noted in the literature as an argument that artists may have been directly attempting to reproduce an American pronunciation model (Trudgill, Gibson “Pop Song English”). Trudgill cites Cliff Richard’s “Bachelor Boy” as an example: the singer, wanting to imitate American rhoticity, incorrectly introduces a linking /r/ in the lyrics ‘You’ll be a bachelor boy…’ realized as [ər ˈbætʃələr ˈbɔɪ], which would not be the case in the targeted accent. In the database, two such cases stand out. In The Housemartins’ song “I’ll Be Your Shelter,” (London 0 Hull 4) singer Paul Heaton from Cheshire, consistently flaps the /t/ in the word ‘shelter’ over the three repetitions of the tokens in the song.8 Although theoretically the phonological environment is suitable to flapping, this /t/ would never actually be flapped in the target American accent, nor is potential voicing indicated in pronunciation dictionaries. Even more striking is the way ABC’s lead singer, Martin Fry, originally from Manchester, pronounces his very own name in a half-spoken part of the song “The Look of Love” (The Lexicon of Love). Though in spoken interviews he introduces himself as [ˈmɑːtɪn], in “The Look of Love,” he can be heard pronouncing it as [ˈmɑːɾɪn]9, effectively modifying the pronunciation of his own name towards what might seem like the targeted American accent – though it is very unlikely an American speaker would flap in this name, most often pronounced with a glottal stop. Gibson hypothesizes that in the early 21st century, as American-influenced pronunciation became increasingly normalized into PSE, this tendency towards hypercorrection decreased. These few examples might suggest that in the 1980s, the process of standardization of PSE is underway, but not yet fully realized.
22Perhaps even more significant as a motivational force behind British style-shift is the cultural authority of American music genres and subgenres. While 1980s British post-punk music is marked by a certain degree of influence from the European electronic music scene or British-branded genres (Straw) such as Madchester, many pop artists nonetheless draw major inspiration from American performers and music genres ranging from rock and roll to rhythm and blues (Frith, “Pop Music”). A closer look into the categorization of music into genres and subgenres is valuable to this analysis given the diversity of pop music productions in the 1980s and might help shed light on different artists’ attitudes towards the US and its influence. For example, artists such as Eric Clapton and Peter Gabriel have both been heavily influenced by the markedly African American genres of blues and soul, respectively; therefore it does not seem surprising that both should flap 100% of their intervocalic /t/ tokens in the database. In an interview, Clapton talks about his sense of kinship with African Americans: “I had to just accept that I was using that identification as a way of strengthening my music […] why else would I have responded to the blues and R&B when I was so young and other music didn't reach me that way?”10 He is also one of the few artists in the database to explicitly mention singing pronunciation, saying “I don't like a lot of the singing styles that come out of England […] I still like people to sing with an American accent.”11 Other singers also acknowledge the influence that America and American-born genres or artists had on them creatively, such as Elton John: “I think America has always been the source of my inspiration, musically anyway,”12 Cliff Richard: “I still say there would have been no Beatles, no Cliff Richard, if it hadn't been for Elvis, Little Richard, Buddy Holly, the Everly Brothers, Chuck Berry… we learned everything from them”13 or Adam Ant: “I think America as a country influenced me immensely.” 14Similarly, it appears genre is an important factor in the case of the band Genesis and its singer Phil Collins’ simultaneous solo career. In the productions of Genesis in the database, Phil Collins’ flapping rate is 68% for 44 tokens, while in his solo productions Collins’ flapping rate is 93% for 41 tokens. Though both rates are well over 50%, the pronunciation discrepancy between Genesis and Phil Collins as a solo artist is statistically significant according to a two-proportion z-test (p < .001). Genesis is a band that belongs to the progressive rock label, a genre that emerged in the UK at the end of the 1960s. Alluding to other bands such as the Rolling Stones “going back to their roots,” guitarist Steve Hackett said in an interview that “this band [Genesis] sounded… sounded [sic] English.”15 On the other hand, Phil Collins’ personal brand of music is very much oriented towards generic pop and may therefore be more easily prone to American influence and the rise of PSE.
23In contrast, a band such as The Jam (14% flapping rate) was known for their distinctly British style and attitude inspired by the mod revival subculture as well as their class-conscious lyrics; elements that make their British identity part and parcel of their music and therefore might explain the reason why they choose not to follow an American-influenced model and stick more closely to their authentic pronunciation. Heaven 17 also state that they “certainly turned down a really large amount of money to not play America,”16 partly because they did not want to play live or tour for extended periods of time. They are coincidentally the only artists in the database to never flap any tokens. When asked about their name, Morrissey of The Smiths (12% flapping rate) said that it is because they “felt it was about time that a pop group identified themselves with normal people, and [they] wanted to be unconventional in rock and roll terms.” Though not as explicit, this declaration can be taken as a rejection of a certain norm in rock and roll, and a wish to engage with a more ordinary audience in “an England that’s fighting very hard to exist.”17 These quotes cannot be taken as definitive arguments, but they do provide examples from both extremes of the word-internal intervocalic /t/ pronunciation continuum which appear to suggest that singers’ attitudes towards their artistic, musical and (to a certain extent) geographical identities may affect their singing pronunciation.
24In spite of these extremes, explained by an overt adoption or rejection of an American-influenced model, a majority of artists in the database are more balanced in their flapping rates – in this case, why only flap on average 50% of the time, when it has been found that American speakers do it well over 90% of the time in environments that allow it (Patterson and Connine)? Conversely, for acts with lower proportions, why flap at all if it is not something that commonly occurs in their speech? A study focusing on British Heavy Metal productions of the same time period found a similar flapping rate (62.1%), though the origins, attitude, and musical inspirations of the two genres differ on all points (Caillol and Ferragne, 2019). This very well may just be linked to the imperfect adoption of a specific singing norm, but there is another, significant parameter in the relationship between song and pronunciation that has rarely been considered and which will be explored in the following subsection: the specificity of the singing voice and its inherent constraints.
25So far, this article has been mainly concerned with sociocultural arguments often found in this type of study, treating style-shift in pop music as an extension of what can normally occur in speech in specific contexts. Though an undeniably essential explanation of the phenomenon, it seems to overlook that this change in pronunciation occurs in a very different context: singing as opposed to speech. Human sound is created through a complex process, starting with the breathing apparatus (consisting of the lungs, the muscles of the abdomen and the diaphragm) producing an over-pressure of air which is transformed into sound by the vocal folds, before passing through the vocal tract resonators which modify this sound into the desired output. While making use of this very same process, the act of speaking and the act of singing differ on all other points. Singing is ruled by its own set of specific constraints, from the need to be on pitch, the necessity to respect the intensity and rhythm of the melody, to the modification of one’s vocal tract to facilitate emission, and more. Although less codified than choral or operatic singing thanks to the use of amplification systems (Sundberg) among other things, pop singing is still very much governed by such restrictions. The distinction between speaking and singing thus introduces to the discussion considerations that are more specifically phonetic in nature and worth examining in order to account for all aspects of style-shift.
26This harks back to Morrissey’s notion of “singability”: the specificity of the singing voice and its constraints has a direct effect on the physical, articulatory setting of the vocal tract, which in turn may be enough to influence the pronunciation of phonemes, namely in the case of word internal intervocalic /t/. According to the Hyper & Hypo-articulation theory, speech, as a motor behavior like any other, tends to “default to a low-cost form of behavior” (Lindblom 413). This applies to speech but may also be particularly true for singing which requires more focus, energy, and precision (Sundberg). Studying the differences between skilled singers and nonsingers, it has indeed been suggested that the ability to minimize effort and still reach a particular goal when singing is “an essential aspect of the beauty of the voice” (Sundberg, What’s So Special About Singers 117). In a questionnaire on the representations of vocal efficiency, singers agree on the importance of production efficiency, in other words, singing in a way which requires little effort (Pillot-Loiseau). An economy of speech (or singing) gestures is therefore an important parameter for singers to control their energy expenditure in a demanding situation. For example, it is well known in singing that vowels are longer in duration and highly favored as they are sonorous and thus ‘carry’ the voice better and more easily than consonants. Combined with the fact that the vocal tract needs to be open and unobstructed to facilitate phonation (Scotto Di Carlo), a plosive consonant such as [t] would be less preferable in singing as well as challenging to produce and sustain since it requires a substantial narrowing of the vocal tract, a stop of airflow and complete closure (Cornut). On the other hand, a flap is usually voiced, meaning “it is simpler, and requires few motor commands to the organs of speech” (Wells 94). Furthermore, it does not require full closure, and is much closer to the vocal tract configuration for vowels (De Jong), which may explain why it is so prevalent in the singing voice, even for British speakers who rarely display this allophone in speech. Morrissey further supports this argument, using the sonority scale as a reference, where the most sonorous speech sounds seem the most favorable for singers. As such, the more sonorous realization of flapped /t/ should be easier to sing than its less sonorous voiceless counterparts. This preference for flapping may also be amplified by the musical parameters of the songs themselves: it may be more difficult to produce a stop in a higher-pitched song, since singing at higher pitches usually implies a greater buccal aperture (Scotto Di Carlo), which in turn does not facilitate the tongue raising, pressing against the sides of the palate, and crashing long enough against the alveolar ridge to create the full, tight closure needed for a plosive. The unconstricted nature of a flap, on the other hand, means its realization should not be too affected by the distance between the tongue and the palate. Although this is purely impressionistic so far, other settings such as rate of articulation or intensity may also influence the articulatory setting of the vocal tract and have an effect on the realization of intervocalic /t/.
27The overarching idea is that rather than just being a deliberate attempt at imitating an American model or adopting a pop singing norm no longer indexed as American, /t/ realization in singing may also depend on the specificity of the singing voice, as well as the musical constraints of a song. The more ‘extreme’ the conditions (higher pitch, greater intensity, faster rate of articulation), the more difficult it might be to produce a well-formed, prototypical stop with full closure. This impression is also partly based on observations made looking at tongue-palate contact during the articulation of stop and flapped /t/ using electropalatography (EPG). EPG is a non-invasive technique used to monitor contact between the tongue and the hard palate during articulation, using a set of electrodes attached to an artificial acrylic palate worn by a speaker. Previous EPG studies have found that flaps require less articulatory control (Recasens) and are associated with a lower degree of tongue constraint (Recasens, "Flapping in unconstrained alveolars"), as opposed to their stop counterpart. Our preliminary work using EPG corroborates the idea that, on average, less contact is needed to produce a flap than a stop. This can be seen in Figure 2, a heatmap representing the roof of the mouth with the alveolar region at the top of the image, and the velar region at the bottom. The color scale represents the average electrode activation in that area during the production of /t/, where red means that tongue contact occurs consistently, and dark blue means no contact has been recorded.
Figure 2. Average contact area for stop and flapped /t/.
28These heatmaps were generated with spoken pronunciation data from a non-native but fluent English speaker who has lived in the United States for five years. Tokens consisted of 15 disyllabics containing an intervocalic /t/ such as metal and city, repeated three times for each /t/ allophone. This pilot study is undeniably limited as it does not concern the singing voice directly, or spontaneous flapping. Nevertheless, it shows differing tongue-palate contact patterns for each allophone. The stop /t/ articulatory pattern shows activation in the alveolar area as required to produce a prototypical stop, as well as lateral tongue bracing on both sides of the palate. This bracing plays a key role in providing tongue stability and building intraoral pressure (Lee et al.) and contributes to the “horseshoe configuration” EPG studies have found speakers produce for alveolar stops (Gibbon et al). The pattern for the flap, on the other hand, shows a lot less overall contact. These preliminary results will require further examination in upcoming experiments, focusing specifically on singing voice pronunciation. They may nevertheless imply that, in specific conditions such as singing, the articulatory pattern of intervocalic /t/ will likely default to one closer to a flap than to a stop. Ultimately, though what we may perceive is rather binary (flap vs stop) – as seen in the auditory analysis where the authors easily reached perfect agreement – what is actually produced might very well be more gradual in nature, varying alongside a pronunciation continuum where multiple factors influence the final realization of /t/. In other words, with degrees of influence varying from one artist to another, style-shift may not just be about sounding American or sticking to a pronunciation standard initially influenced by an American accent, but also has to do with the fact that some features which happen to be American are better suited to singing.
29This paper attempted to look at the now widely recognized use of an American-inspired Pop Song English pronunciation norm, through the example of the realization of word internal intervocalic /t/ in the sung productions of the most successful British artists of the 1980s. An extensive quantitative analysis of nearly fifty artists’ pronunciation of this feature was provided, showing that despite never flapping in speech, a majority of artists tend to flap most of the time when singing. Certain sociocultural factors were presented as an explanation for this style-shift, evidenced by artists’ own perspective on the matter. The rest of the article aimed to develop an additional, complementary justification for the use of /t/ flapping specifically and PSE in general looking more closely at phonetic and articulatory factors. What is now a PSE standard no doubt originated from some kind of American influence, but there is a further phonetic argument why this type of pronunciation established itself as a norm. Factors pertaining to the specificity of the singing voice and its articulatory constraints, at least in the case of intervocalic /t/ pronunciation, may also have played a part in the widespread adoption of the flapped realization in singing.
30List of artists and albums featured in the database18.
Artist
|
Album (year released)
|
Song
|
ABC
|
The Look of Love (1982)
|
All of my Heart; Poison Arrow; The Look of Love
|
Adam and the Ants
|
Kings of the Wild Frontier (1980)
|
Don’t Be Square (Be There); Feed Me to the Lions; Jolly Roger; Kings of the Wild Frontier; Los Rancheros
|
Prince Charming (1981)
|
Stand and Deliver; That Voodoo!
|
Brian Ferry
|
Boys and Girls (1986)
|
Don’t Stop the Dance; Sensation
|
Bros
|
Push (1988)
|
Drop the Boy; Liar; Love to Hate You; Shocked; Ten out of Ten; When Will I Be Famous?
|
Cliff Richard
|
Wired for Sound (1985)
|
Better Than I Know Myself; Broken Doll; Oh No, Don’t Let Go; Summer Rain; Wired for Sound
|
Always Guaranteed (1987)
|
Once Upon a Time; One Night; Two Hearts
|
Stronger (1989)
|
Clear Blue Skies; I Just Don’t Have the Heart; Joanna; The Best of Me
|
Culture Club
|
Colour by Numbers (1983)
|
Changing Every Day; Church of the Poison Mind; It’s a Miracle; Miss Me Blind; Mister Man
|
Curiosity Killed the Cat
|
Keep your Distance (1987)
|
Down to Earth; Free; Ordinary Day
|
David Bowie
|
Scary Monsters (and Super Creeps) (1980)
|
Ashes To Ashes; Scary Monsters (and Super Creeps); Scream Like a Baby; Teenage Wildlife; Up The Hill Backwards
|
Let’s Dance (1983)
|
China Girl; Criminal World; Let’s Dance; Shake It
|
Depeche Mode
|
Violator (1990)
|
Blue Dress; Clean; Enjoy the Silence
Sweetest Perfection; World in My Eyes
|
Dexys Midnight Runners
|
Too Rye Ay (1982)
|
I’ll Show You; Jackie Wilson Said (I’m In Heaven When You Smile)
|
Duran Duran
|
Duran Duran (1981)
|
Anyone Out There; Careless Memories; Friends of Mine; Sound of Thunder; To the Shore
|
Rio (1982)
|
Hungry Like the Wolf; My Own Way; The Chauffeur
|
Seven and the Ragged Tiger (1983)
|
New Moon on Monday; Of Crime and Passion; The Seventh Stranger; Union of the Snake
|
Electric Light Orchestra
|
Time (1981)
|
From the End of the World; Here is the News; Hold on Tight; The Lights Go Down; Ticket To The Moon; Twilight
|
Elton John
|
Too Low for Zero (1983)
|
Cold As Christmas (In the Middle of the Year); I’m Still Standing; One More Arrow; Religion; Saint; Whipping Boy
|
Breaking Hearts (1984)
|
Breaking Hearts Ain’t What It Used To Be; Burning Buildings; Did He Shoot Her?; Lil’ ‘Frigerator; Restless; Sad Songs (Say So Much)
|
Ice on Fire (1985)
|
Candy By the Pound; Cry to Heaven; Nikita; Tell Me What the Papers Say; This Town; Wrap Her Up
|
Sleeping with the Past (1989)
|
Amazes Me; Blue Avenue; Club At the End of the Street; Durban Deep; Healing Hands; Sleeping With the Past; Stone’s Throw from Hurtin’
|
Erasure
|
The Circus (1987)
|
Don’t Dance; It Doesn’t Have To Be
|
The Innocents (1988)
|
A Little Respect; Hallowed Ground; Imagination; Ship of Fools; Yahoo!
|
Wild (1989)
|
Blue Savannah; Brother and Sister; Drama!; How Many Times?; Piano Song; Star
|
Eric Clapton
|
Journeyman (1989)
|
Hard Times; Lead Me On; Old Love; Running On Faith
|
Fine Young Cannibals
|
The Raw and the Cooked (1989)
|
It’s Ok (It’s Alright)
|
Frankie Goes to Hollywood
|
Welcome to the Pleasuredome (1984)
|
Born To Run; Snatch of Fury (Stay); Relax; San Jose (The Way); The Power of Love; War
|
Genesis
|
Duke (1980)
|
Alone Tonight; Behind The Lines; Cul de Sac; Duke’s Travel; Man Of Our Times; Misunderstanding; Open Door; Turn It On Again
|
Abacab (1981)
|
Another Record; Keep It Dark; Like It or Not; Me and Sarah Jane
|
Genesis (1983)
|
Just a Job to Do
|
Invisible Touch (1986)
|
Anything She Does; In Too Deep; Invisible Touch
|
George Michael
|
Faith (1987)
|
Father Figure; Monkey
|
Listen without Prejudice (1990)
|
Cowboys and Angels; Freedom! ’90; Mothers Pride; Praying Time; They Won’t Go When I Go; Waiting for That Day
|
Go West
|
Go West (1985)
|
Call Me; Eye to Eye; Goodbye Girl; Haunted; Missing Persons
|
Haircut One Hundred
|
Pelican West (1985)
|
Calling Captain Autumn; Fantastic Day; Kingsize (You’re My Little Steam Whistle); Love Plus One; Marine Boy; Milk Film
|
Heaven 17
|
The Luxury Gap (1983)
|
Crushed By the Wheels of Industry; The Best Kept Secret; Who’ll Stop the Rain
|
Johnny Hates Jazz
|
Turn Back the Clock (1988)
|
What Other Reason
|
Level 42
|
World Machine (1985)
|
Coup D’Etat; Dream Crazy; It’s Not the Same For Us
|
Running in the Family (1987)
|
Children Say; It’s Over; The Sleepwalkers; Two Solitudes (Everyone’s Love In The Air)
|
Nik Kershaw
|
Human Racing (1984)
|
Bogart
|
The Riddle (1984)
|
City of Angels; Wild Horses
|
Orchestral Manoeuvres in the Dark
|
Architecture and Morality (1981)
|
She’s Leaving; Souvenir; The New Stone Age
|
Paul McCartney
|
Tug of War (1982)
|
Ballroom Dancing; Dress Me Up as a Robber; Ebony and Ivory; Here Today; Somebody Who Cares; Wanderlust
|
Paul Young
|
No Parlez (1983)
|
Ku Ku Kurama
|
The Secret of Association (1985)
|
I’m Gonna Tear Your Playhouse Down; Man in the Iron Mask; Soldier’s Things
|
Pet Shop Boys
|
Please (1986)
|
I want a Lover; Tonight is Forever; Two Divided by Zero; Violence; West End Girls
|
Actually (1987)
|
Hit Music; One More Chance; Shopping
|
Introspective (1988)
|
Always On My Mind / In My House; Domino Dancing; I Want a Dog
|
Peter Gabriel
|
So (1986)
|
Mercy Street; That Voice Again
|
Phil Collins
|
Face Value (1981)
|
I Missed Again; If Leaving Me Is Easy; I’m Not Moving; In the Air Tonight; This Must be Love; Tomorrow Never Knows
|
No Jacket Required (1983)
|
Don’t Lose My Number; Inside Out; One More Night; Who Said I Would
|
…But Seriously (1989)
|
All of my Life; Another Day in Paradise; Colours; Do You Remember?; Father to Son; Hang in Long Enough; I Wish it Would Rain Down; That’s Just the Way It Is
|
Pink Floyd
|
The Wall (1979)
|
Another Brick in The Wall; Comfortably Numb; One of My Turns; The Thin Ice; Waiting For The Worms; Young Lust
|
Rick Astley
|
Whenever You Need Someone (1987)
|
Never Gonna Give You Up; The Love Has Gone; When I Fall In Love
|
Roxy Music
|
Flesh and Blood (1980)
|
Oh Yeah!; Over You; Running Wild: Same Old Scene
|
Avalon (1982)
|
Avalon; More Than This; Take A Chance With Me; To Turn You On
|
Simply Red
|
Picture Book (1985)
|
Jericho; Money’s Too Tight (to Mention); No Direction; Open Up the Red Box; Picture Book
|
Men and Women (1987)
|
Ev’ry Time We Say Goodbye; I Won’t Feel Bad; Let Me Have It All; Love Fire; Move On Out; Shine; Suffer; The Right Thing
|
A New Flame (1989)
|
More; She’ll Have to Go
|
Soft Cell
|
Non Stop Erotic Cabaret (1982)
|
Bedsitter; Chips On My Shoulder; Entertain Me; Frustration
|
Spandau Ballet
|
True (1983)
|
Foundation; Heaven Is a Secret; True
|
Parade (1984)
|
Highly Strung; I’ll Fly For You; Only When You Leave
|
Sting
|
The Dream of the Blue Turtles (1985)
|
Consider Me Gone; Love is the Seventh Wave; Russians
|
Tears for Fears
|
The Hurting (1983)
|
Change; Mad World; Start of the Breakdown; The Prisoner; Watch Me Bleed
|
Songs From the Big Chair (1985)
|
Listen
|
The Seeds of Love (1989)
|
Swords and Knives
|
The Beautiful South
|
Welcome to the Beautiful South (1989)
|
I’ll Sail this Ship Alone; Woman in the Wall
|
Choke (1990)
|
I’ve Come For my Award; Let Love Speak Up Itself; My Book
|
The Christians
|
The Christians (1988)
|
Born Again; Ideal World; One in a Million; Sad Songs; When the Fingers Point; Why Waltz
|
The Housemartins
|
London 0 Hull 4 (1986)
|
Get Up Off Our Knees; Happy Hour; He Ain’t Heavy, He’s My Brother; I’ll Be Your Shelter (Just Like a Shelter); Lean On Me; Over There; People Get Ready; Sheep
|
The Human League
|
Dare! (1981)
|
Do or Die; Don’t You Want Me; Love Action (I Believe in Love); Open Your Heart; Seconds; The Sound of the Crowd
|
The Jam
|
The Gift (1982)
|
Just Who Is the 5 O’Clock Hero?; Precious; Running on the Spot; The Gift; The Planner’s Dream Goes Wrong; Trans-Global Express
|
The Police
|
Outlandos d’Amour (1978)
|
Next To You; Roxanne
|
Reggatta De Blanc (1979)
|
Contact; No Time This Time; Walking on the Moon
|
Zenyatta Mondatta (1980)
|
Bombs Away; Canary in a Coalmine; Don’t Stand So Close to Me; Driven To Tears; When the World Is Running Down
|
Ghost in the Machine (1981)
|
Every Little Thing She Does Is Magic; Rehumanize Yourself
|
Synchronicity (1983)
|
Every Breath You Take; King of Pain; O My God; Synchronicity II; Walking in Your Footsteps; Wrapped Around Your Finger
|
The Rolling Stones
|
Emotional Rescue (1980)
|
All About You; Down in The Hole; Send It to Me; She’s So Cold; Summer Romance
|
The Smiths
|
The Smiths (1984)
|
Hand In Glove; Miserable Lie; Reel Around the Fountain; Still Ill; What Difference Does It Make; You’ve Got Everything Now
|
The Stone Roses
|
The Stone Roses (1989)
|
Bye Bye Badman; Don’t Stop; I Am the Resurrection; (Song for My) Sugar Spun Sister; Waterfall
|
The Style Council
|
Café Bleu (1984)
|
A Gospel; Headstart for Happiness; Here’s One That Got Away; My Ever-Changing Moods; The Paris Match; The Whole Point Of No Return; You’re The Best Thing
|
Thompson Twins
|
Into the Gap (1984)
|
Hold Me Now; Sister of Mercy; Who Can Stop the Rain; You Take Me Up
|
Wham!
|
Fantastic (1983)
|
Bad Boys; Club Tropicana; Come On!; Love Machine
|
Make it Big (1984)
|
Careless Whisper; Everything She Wants; Freedom; If You Were There
|