An Acoustic Analysis of Vowels
In this essay, I investigate changes in the vowel system across time in the community of South-East England. I have collected and analysed data from four speakers in this community (two younger and two older). In section 1.0, I outline the method I used to collect this data and why I went about it in this way. In section 2.0, I present my data in vowel plots. In section 3.0, I discuss my findings and possible reasons for the data, which I support with evidence from other sources. In this section, I specifically look at the vowels [ɪ, e, ɛ, ɒ, ʌ, ʊ, uː]. Finally, in section 4.0, I conclude what my data and supporting evidence say about changes in the vowel system over time in this community.
1.0 Methodology
In my study, I used four participants: two older female and male (OF & OM) participants and two younger female and male (YF & YM) participants. According to Wells (1982), once past puberty, people’s speech is not as affected by ongoing sound changes; thus, children show the most noticeable differences in changes to the sound system. The National Health Service ([NHS], 2019) states that the average age of puberty in girls is 11 years old, whilst for boys it is 12 years old. Therefore, I chose my younger participants to be below these averages; the YF was 6 years old, and the YM was 8 years old. To leave a large age gap to allow for sound change, I ensured 50 years difference at least; the OM was 67 years old, and the OF was 59 years old. I used an equal number of males and females in each category to reduce confusion between frequencies due to sex differences, as according to Reetz and Jongman (2008), males and females often produce different formant frequencies due to vocal tract shape differences. Moreover, all participants were native speakers of British English. As stated by Aronoff & Rees-Miller (2003), the majority of non-native speakers of a language will be distinguishable from native speakers by their accent. Therefore, by having native British English speakers, I removed the effects of non-native pronunciation differences affecting the data.
The participants were asked to record themselves reading a couple of sentences from a well-known children’s story:
Once upon a time, there was a little girl who lived in a village near the woods. She always wore a red riding cloak, so everyone called her Little Red Riding Hood. (Scribe Bee, 2019)
I chose a children’s story as the reading task for several reasons. Firstly, as my youngest participant was six years old, I needed reading material that they would be able to read without the aid of an adult. I chose a well-known children’s story as the vocabulary would be both readable and recognisable to younger and older participants alike. Secondly, I chose this reading task to record a style of speech production most likely to come about in everyday life. As stated by Wells (1982), there are differences in pronunciation between our casual and formal styles of speech. A very formal style of speech is elicited when asking people to read a list of words; this is made more so when they are to be recorded (Wells, 1982). By asking participants to read a few sentences from a well-known story, I decreased the effects of this formal style. In addition, by asking participants to record themselves in their own homes, the effects of the presence of an interviewer were reduced, further enforcing the atmosphere of an everyday reading task.
From the reading task, I identified the stressed monophthongs in some of the words (above in bold), which were of interest, which in RP would be [ʌ, ɒ, u, ɪ, ʊ, ɔ, e, ɛ]. Measurements of the F1 and F2 frequencies for each vowel were measured in Praat and put into vowel plots. The measurements for each vowel are labelled by their appropriate lexical set in 2.0. However, it must be understood that the data may have been affected by the environmental context of the vowels. To counter this, I ensured I extracted formant frequencies from the most central parts of each vowel to limit the coarticulatory effects of the preceding and following consonants. Although a larger study would benefit from both set contextual environments (such as hVd) and an everyday reading task.
2.0 Data
Figure 1: Vowel Formant Values for All Speakers non-normalized
Figure 2: Older Male Vowel Formant Values non-normalized
Figure 3: Older Female Vowel Formant Values non-normalized
Figure 4: Younger Male Vowel Formant Values non-normalized
Figure 5: Younger Female Vowel Formant Values non-normalized
3.0 Discussion
Figure 1 shows a vowel plot presenting data collected from all four participants, the OM (red), OF (blue), YM (green), and YF (orange). Overall, the data formant frequencies collected varied widely between the participants due to individual differences such as vocal tract size and shape. As shown in Figure 1, the OM speaker has the lowest frequencies for most of the vowels recorded. As the formant frequencies correspond to the resonance frequencies of the vocal tract, they are affected by vocal tract size and shape (Reetz and Jongman, 2008). Adult males typically have larger vocal tracts than women and children (Wells, 1982); men also have longer vocal folds, which results in lower frequencies (Reetz and Jongman, 2008). Therefore, data between speakers will be compared with respect to their individual vowel spaces and not their exact formant frequencies.
3.1 Goose-fronting
As shown in the vowel formants in 2.0, the younger speakers have a more fronted [uː] vowel (labelled ‘GOOSE’) than the older speakers, shown by their larger F2 frequencies. As stated by Hawkins & Midgley (2005), the fronting of [uː] is well established amongst younger speakers of RP (those born during the 21st century). However, the older speakers in my data also produce a more fronted [uː] using a vowel more like [ʉː]. This is not surprising, as cited in Hawkins and Midgley (2005: 184), Wells (1982: 145) argued that [uː] is more centralised in mainstream RP and Urban English than old RP.
In the younger speakers, the vowel is not just fronted but lowered, making it more likely to be that of the rounded [Y]. Stuart-Smith et al (2017, cited by Lawson, Smith and Rodger, 2019: 4367) demonstrate that such ‘Goose lowering’ has been recorded in studies on Scottish dialects of English.
3.2 ʊ Fronting
In 2.0, the vowel plots show that the [ʊ] vowel in WOOD is more fronted in the younger speakers (with respect to their individual vowel spaces) than in the older speakers. As reported in Hawkins and Midgley (2005:184), Wells (1982:145) described [ʊ] in RP as more likely now to be unrounded and centralised. A study by Kerswill and Williams (2000) supports this, stating that the unrounding, lowering and fronting of [ʊ] to a [ə] like vowel was ongoing. This explains the fronting (higher F2 values) and lowering (increased F1 values) in the younger speakers to the older speakers. Therefore, it is not surprising that the younger speakers have productions of [ʊ] much closer to schwa, and the older speakers are much closer to the original [ʊ].
It should be noted that despite [ʊ] fronting and lowering in the younger speakers, all speakers still retain a strong distinction between the vowels [ʊ] and [ʌ]. With all speakers keeping [ʌ] with high F1 and low to middle F2 frequencies, whilst the F1 frequency for [ʊ] is always much lower. Jansen and Braber (2021:767) state that this is distinctive of southern British English and distinguishes this community from the East Midlands variety.
3.3 Lot
As can be seen in the vowel plots, the LOT vowel [ɒ] (the word ‘was’ in the reading) shows large variation between the older and younger speakers. In the older speakers, the vowel has a low F2 and high F1, thus the vowel appears where it would be expected in the vowel quadrilateral. However, the younger speakers display a much higher and more fronted vowel (becoming more centralised) that sounds more like [ə]. As reported in Kamata (2008: 60), Torgersen and Kerswill (2004: 40) state that the raising of LOT is an ongoing process in British English spoken in Ashford, London.
Kerswill (2003:223, cited by Beal, 2010: 73) states that dialect levelling can occur because of geographical diffusion, a process in which a community’s dialect features spread to surrounding communities that are less economically or culturally dominant. Therefore, it makes sense that the younger speakers in this county of Hertfordshire, Southeast England, are being influenced by its larger and more prosperous neighbouring city of London.
3.4 DRESS
In the vowel plots, the younger speakers have a more lowered vowel in DRESS than the older speakers; younger speakers produce a vowel more like [æ], and the older speakers produce vowels like [ɛ]. According to Hawkins and Midgley (2005), an increasing F1 in the monophthong [ɛ] is becoming increasingly common, as well as the vowel’s perception of sounding more open. Hawkins and Midgley (2005) found that speakers born before 1991 will likely have markedly lower F1 frequencies than those born after this date, as [ɛ] had begun to fall in the speech of younger speakers from 1981 but showed noticeable change 10 years after this. This explains the lack of lowering in the results of my older speakers’ data, as both were born before 1991; the OM was born in 1955, and the OF was born in 1968.
3.5 KIT
In the vowel plots, it can be observed that the younger speakers have more lowered versions of [ɪ] in the KIT lexical set than the older speakers. As reported in Kamata (2008:60), Trudgill (2004: 42-43) argues the vowel in KIT is a part of the drag chain shift in the South-east of England and involves three stages: the lowering of the vowel in TRAP, then the vowel in DRESS, then subsequently KIT. Wells (1982, cited in Hawkins and Hidgley 2005:184) states that [ɪ] is becoming more open and central through this ongoing change. However, despite the apparent lowering of the vowel, neither of the younger speakers demonstrates much centralising of the vowel as their F2 frequencies remain high. With respect to the F2 frequency of their FLEECE vowel, the
F2 frequency of their [ɪ] vowel only reduces by 208 Hz for the YM and 333Hz for the YF. Therefore, it is arguable that for this community, the drag chain shift is in effect, but no centralisation through tongue backing is occurring.
3.6 Strut
As shown in the vowel plots, all speakers demonstrate a largely lowered [ʌ] vowel and slight fronting. As cited in Kamata (2008: 60), Trudgill (2004: 113-4) states the fronting and lowering of [ʌ] is an ongoing process which is most progressed in Southeast England and London. As this is a change that has already progressed, it explains the fronting in all speakers. However, the speaker who demonstrates the most fronting is the YF. According to Wells (1982), once past adolescence, a person’s accent does not change much and thus marked changes in accent caused by ongoing processes are more likely to be seen in those who have not reached puberty. As the youngest, the YF is the most likely to demonstrate this ongoing fronting, and this is shown in Figure
3.7 Anticlockwise Chain shift
As previously described, the data collected in this study shows a shifting of the vowels [ɪ, e, ɛ, ɑ, ʊ uː] anticlockwise around the vowel space. According to Hawkins and Midgley (2005), there is a current tendency for the anticlockwise rotation of monophthongal vowels on the outside of the vowel quadrilateral, with particularly evident changes in [ɛ, æ, ʊ uː]. They also state that the position of the vowel [iː] remains stable and that [ɑ, ɔː] do not raise or move back very much.
In all the vowel plots, the location of the high front vowel [iː] in FLEECE remains consistent in its location, retaining the low F1 and high F2 frequencies characteristic of [iː]. The vowel in NORTH also remains in a position consistent with the expected formant frequencies of [ɔː]. However, my data found that there is a noticeable raising of the LOT vowel for the younger speakers as well as an increase in F2 for the YF. This is not consistent with Hawkins & Midgley’s (2005) findings, which leads to the possible conclusion that this is a vowel change occurring in standard English in the South-East of England and not a change fully revealing itself in RP, which they were measuring.
The fronting of the STRUT vowel in my data is also not supported by Hawkins and Midgley’s (2005) results. However, as previously mentioned, support for /ʌ/ fronting is provided by Trudgill (2004: 113-4, cited in Kamata, 2008: 60). This suggests that this fronting might also be more distinct for this community than for RP speakers.
4.0 Conclusion
It can be concluded from my data that vowel system change has occurred in the Southeast of England over the last 50 years. Specifically, the vowel system of younger speakers is showing the anticlockwise chain shift of [ɪ, e, ɛ, ɑ, ʊ uː] and the fronting of [ʌ]. However, [iː] and [ɔː] seem not to have changed with time.
Despite the evidence presented in 2.0, it could be argued that the similarities between the younger speakers could be attributed to not just an age difference between them and the older speakers but to the fact that they are siblings living in the same household. However, Feiser’s (2009) study found weaker similarities in F1-4 frequencies between siblings than between non-related speakers. This provides support to the data demonstrating age-related differences in vowel production as opposed to a reflection on shared living spaces.
References
Aronoff, M., & Rees-Miller, J. (Eds.). (2003). The Handbook of Linguistics [Ebook]. Oxford: John Wiley & Sons, Incorporated. Retrieved from https://ebookcentral.proquest.com/lib/lancaster/detail.action?docID=351555.
Beal, J. (2010). An Introduction to Regional Englishes : Dialect Variation in England [Ebook]. Edinburgh: Edinburgh University Press. Retrieved from https://ebookcentral.proquest.com/lib/lancaster/detail.action?docID=647698.
Feiser, H. S. (2009). Acoustic similarities and differences in the voices of same-sex siblings. In 18th Annual Conference of the International Association for Forensic Phonetics and Acoustics (IAFPA), Cambridge, UK.
Hawkins, S., & Midgley, J. (2005). Formant frequencies of RP monophthongs in four age groups of speakers. Journal Of The International Phonetic Association, 35(2), 183-199. doi: 10.1017/s0025100305002124
JANSEN, S., & BRABER, N. (2020). <scp>foot</scp>-fronting and <scp>foot</scp>–<scp>strut</scp> splitting: vowel variation in the East Midlands. English Language And Linguistics, 25(4), 767-797. doi: 10.1017/s1360674320000325
Kamata, M. (2008). An acoustic sociophonetic study of three London vowels. (Ph.D). University of Leeds.
KERSWILL, P., & WILLIAMS, A. (2000). Creating a New Town koine: Children and language change in Milton Keynes. Language In Society, 29(1), 65-115. doi: 10.1017/s0047404500001020
Kerswill, Paul. (2003) Dialect levelling and geographical diffusion in Birisih English. In David Britain & Jenny Cheshire (Eds.), Social Dialectology: In Honour of Peter Trudgill, (pp. 223-45). Retrieved from https://ebookcentral.proquest.com/lib/lancaster/reader.action?docID=623115
Lawson, E., Stuart-Smith, J., & Rodger, L. (2019). A comparison of acoustic and articulatory parameters for the GOOSE vowel across British Isles Englishes. The Journal Of The Acoustical Society Of America, 146(6), 4363-4381. doi: 10.1121/1.5139215
National Health Service. (2019). Early or delayed puberty. Retrieved 24 April 2022, from https://www.nhs.uk/conditions/early-or-delayed-puberty/
Reetz, H., & Jongman, A. (2008). Phonetics : Transcription, Production, Acoustics, and Perception [Ebook]. Chichester: John Wiley & Sons, Incorporated. Retrieved from https://ebookcentral.proquest.com/lib/lancaster/detail.action?docID=819405.
Bee, S. (2019, May 29). Little Red Reding Hood. Retrieved from medium.com: https://medium.com/@scribebeee/little-red-riding-hood-fabf1f9d74ad
Stuart-Smith, J., José, B., Rathcke, T. V., Macdonald, R., & Lawson, E. (2017). Changing sounds in a changing city: an acoustic phonetic investigation of real-time change over a century of Glaswegian.
Torgersen, E., & Kerswill, P. (2004). Internal and external motivation in phonetic change: Dialect levelling outcomes for an English vowel shift. Journal Of Sociolinguistics, 8(1), 23-53. doi: 10.1111/j.1467-9841.2004.00250.x
Wells, J. (1982). Accents of English: Volume 1 [Ebook]. Cambridge: Cambridge University Press. Retrieved from https://books.google.co.uk/books?hl=en&lr=&id=Ty5RoXyTKQsC&oi=fnd&pg=PR15& ots=RjP1i06VFL&sig=mPMftEbQWbgjazb9p5uCQtpOV4&redir_esc=y#v=onepage&q&f=false.
Waygaze 
Leave a comment