ftfitcd}g#lie?gKJvJS(ISFeigiYFfiIRk eg47e 5‑‑10 (1993)
THE INITIAL
OF
POSITION OF THE FIRST ELEMENT /at/ AND /au/ IN ENGLISH
Yutaka SOEDA
I
J.D. O'Connor has the following statement:
/ai/ glides from A/ to /i/, and the loudness becomes less as the glide progress‑
es. Say fAn fun, and then add /i/ after the !IA/, with a smooth glide; this will give you fain fine. ...For /au/ start with /A/. Say tAn ton, and then after the /A/‑sound add an /u/; this should give taun lown.'
He seems to identify /a/ with /A/ here. Our first reaction toward his remark would be either a) that there may be a phonetic reason for it or e) that there may be an educational consideration behind it ; that is, his instructions may be a pedagogical device, nothing more. As to /ai/, Roach agrees with him that ̀this diphthong begins with an open vowel which is between front and back ; it is quite similar to the A of the words ̀cut', ̀bun'.2 However, Roach's description of /au/
is quite different ; ̀. . . this diphthong begins with a vowel similar to a: but a little more front.'3
The purpose of this paper is to examine and confirm the nature of the first element of /ai/ and /au/, i. e. the difference, if any, between the two, based on the articulatory and acoustic data available to us.
and
To facilitate matters described by some
A comparison
Bronstein Gimson Jones Kenyon Ladefoged O'Connor
ll
,let us first see how the two diphthongs are transcribed of the major phoneticians, both British and American.
Table 1
of some systems for transcribing /ai/ and /au/
ai au (1984), au (1988)
ai (1949), ai (1960) au (1949), au (1960)
ai au (1988), au (1990)
Roach ai au Ward ai au
Wells ai au (1982), au (1990)
Wise ai au
Descriptions :
Bronstein argues that the initial position of /ai/ varies between [a] and
[a] .4
Gimson's comment is as follows : ̀In the case of/ai, au/, there may be a common (open central) staning point,...for many speakers, the lst element of /ai/
and /au/ may in fact be identical.'5 (There seems to be no particular reason for the replacement of [au] (1984) by [au] (1988) except for bringing the transcription in line with the English Pronouncing Dictiona?y.)
Ladefoged remarks that both of the diphthongs [ai, ao] as in "high, how"
start from more or less the same low central vowel position, midway between [Ee] and [a] and slightly closer to [A] than to any of the other vowels.6 (This reminds us of O'Connor's instructions quoted earlier.) But he also says, "The diphthong [au] in ̀how' usually starts with a slightly more front quality than that in ̀high.'7
Ward observes that this (i.e. the starting point of the first element of ai) varies from ee to a... .8 au might well be written as au, since the first element lies between a and a.9
Wells' comment on [ai] is that phonetically, it is a wide diphthong with a starting‑point which is open, unrounded, and most usually centralized‑front, [ai], though front and central variants, [ai･x‑a+i] are also common within the standard accents.iO ...phonetically, [au] is a wide diphthong with a starting‑point which is open, unrounded, and most usually central, about halfway between cardinals [a] and [a]; though centralized‑front and centralized‑back varianls, [a‑u‑‑a+u], are also common within the standard accents.ii (italics mine) (Like Gimson and O'Connor, Wells replaced [au] (1982) by [au] (1990) only to bring the transcription in line with the English Pronouncing Dictionary.)
Wise introduces us to various transcriptions which are used in phonetic literature to express the diphthong [ai], such as [ai], [ai], [ae], [eei], [eee], [aej], [ai], [ay], [aj], [aj], [Qj], [Ai], [Ae], [oe], and [ae].i2
In summary, then, all of them agree that considerable latitude is permitted between the values [al and [a] for the first element of both /ai/ and /au/, with the possibility of being identical, but also with the greater possibility of being different, as shown in the transcription [au] used by some of them.
THE INITIAL POSITION OF THE FIRST ELEMENT OF /ai/ AND /au/ IN ENGLISH 7
Unlike consonants, English vowels in particular have a lot of variants. Thus the starting point of/ai/ and /au/ is no exception. The reason why both diphthongs tend to attract so much attention is clearly not only because their initial position can or cannot be identical, but also because they are unique in forming a pair of diphthongs, so to speak. Although diphthongs are usually divided into three groups in terms of how to end the second element, i.e. ending in [ot [i] and [u], little, if any, attention has been paid to the fact that /ai/ and/au/ make a pair of diphthongs in terms of having the first element in common.
m In order to be able to offer some visual use vowel charts used in describing English.
materials, it will be convenlent to
e
Fig. 1. ‑/A/ (Roach l991) Fig. 3. ‑/a:/ (Roach 199l)
Fig. 2. ‑/ai/ (Roach 1991) Fig. 4. ‑/au/ (Roach I991)
As we can see from Figs.1 and 2, the first element of /lx/, and, from Figs. 3 and 4, that of /au/ is close to /a:/.
to tell the whole story, because those four figures do consideration at all.
/ai/ is certainly close to But that does not seem not take variants into
8
i
e
Somerset Tyneside
E
Devon RP
a
u
o
Yutaka SOEDA
i
erkse ales
Somerset E
a
Cockney
u
o
oo
B'ham
aCockney aN. Ireland
RP
Fig. 5. ‑/ai/ variants (O'Connor i988) Fig. 6. ‑/au/ variants (O'Connor l988)
The starting point may vary a good deal as shown in Figs. 5 and 6. We can see from these that in most kinds of English the first element has a central quality intermediate between front and back, and is not significantly different fot /ai/
and /au/, and that RP has a back starting point for /au/ near Cardinal 5 [a]
and a front starting point for/ai/ near Cardinal 4 [al. In so far as RP is concerned, Figs. 7 and 8 give more practical and useful information on this matter.
i: '‑
I‑‑‑ ‑‑
3:,e‑‑ ‑ e‑‑‑
ee ‑‑‑
‑'
‑‑u:
‑‑ U
‑‑ ‑‑ O:
‑‑‑ o
el ‑‑‑ ‑‑‑ eu
‑‑ OLI
‑‑t l
‑
A a: al, au
Fig. 7. ‑RP monophthongs: acceptable are‑ Fig. 8. ‑RP falling diphthongs: acceptableas. (Gimson l989) onset areas. (Gimson 1989)
As shown in Fig. 8 in particular, the initial position of the first element of /ai/
and /au/ can be identical in RP, sharing a common acceptable area.
Finally let us look at some acoustic data on the subject. According to Shimoda (1981), who made an acoustic experiment on /ai/ and /au/ in terms of formant frequencies, it was acoustically proved that the first element of /ai/ is a little closer to /ee/ than that of /au/, which is a bit closer to /a:/ than that of the former, and that neither of them are the same as /ee/ or /a:/.
THE INITIAL POSITION OF THE FIRST ELEMENT OF /ai/ AND /au/ IN ENGLISH 9
200 250
""v‑‑rsL JU s
soo M
ai au
(Fl)
3K 2K IK 500 400 low
Fig. 9. ‑First and second formants for Fig. 10. ‑A combined acoustic and auditory rep‑IKthediphthongs.(Borden&Harris resentation of some of the vowels of
l983) American English. (Ladefoged 1982)
In Fig. 9 the first formant frequencies are plotted on the vertical axis and the second formant frequencies on the horizontal axis. Notice the direction of formant movement shown by the arrows; the arrows for /ai/ and /au/ cross each other at the bottom. Fig. 10 also shows much the same formant movement. Their crossing at the bottom clearly indicates that in some cases the first element of /ai/ can be closer to /a/ than that of /au/ and that the first element of /au/
can be closer to /ee/ than that of /ai/ in some other cases.
By and large, Figs. 1 through 8, based on articulatory (or auditory) data, and Figs. 9 and 10, based on acoustic data, seem to show almost the same features of the initial position of the first element of /ai/ and /au/ within the standard accents of English.
rv
This may have seemed a lot of fuss about a trivial problem, but it does, as most phoneticians point out, have considerable interest, both articulatory and acoustic. In considering the problem, which is twofold, we cannot help focussing on two phonetic aspects of it at the same time;variants of each /al sound and, in addition, a comparison of the first element of /ai/ with that of /au/. In order to show the starting point for /ai/ and /au/ in RP, we can use any of [a], [A], and [a] . So O'Connor's instructions quoted at the very beginning are a phonetically acceptable suggestion as well as a pedagogical device. But, unfortunately, it is absolutely impossible for any sign to represent all variants. In practice, then, it would be wise to regard the initial position of /ai/ and /au/ as identical in many cases and not to forget that it can be different from dialect to dialect and
(F2‑Fl) high
ooooooo
N‑
8￡ gN o8T‑1 89 8a 8oo 8ts
oooo
on oooowco Hz
'i
.IU eu 300
el 400
Oe eo
front e.
eo
500 3I
‑E
e! o‑ 600
ee ea 700
10 Yutaka SOEDA even from person to person．
Notes
1
2
3 4 5 6 7 8 9
10 11 12
：
0 Connor（1990）， P．85．
Roach（1991）， p．21．
Ibid．， p．22．
Bronstein（1960）， p．195．
Gimson（1989）， p．137．
Ladefoged（1982）， p．76．
Ibid．， p．77．
Ward（1972）， p．116．
Ibid．， p．118．
Wells（1982）， p．149．
Ibid．， p．151．
Wise（1957）， p．103．
References
Borden， G． J．＆K． S Harris．1983． S勿碗56加6θP7勿zθ飢（2nd edn）Williams＆Wilkins．
Bronstein， A． J．1960．丁勿P％oπ襯6ガα渉∫oπqん、4〃z67♂6αηEπg1ガε乃。 Appleton−Century−Crofts， Inc．
Catford， J． C．1988．ノ1 P7α6だαzl Iη〃04zκガ。π o P乃。η6琵6s． Clarendon press， Oxford．
Clark， J．＆C． Yallop．1990．．∠1η1π〃04π6 ∫oη 01）ぬ。彫 ゴos（島一P加η010gy． Basil BlackwelL Fry， D． B．1980．：τ「加ρ勿sゴos（ゾSヵθθ碗． Cambridge University Press．
Gimson， A． C．1975．．4 Pη漉6α1 Co％zsθqプEπg1納P 〃魏。謝ゼ。π． Edward Arnold．
Gimson， A． C．1989．！肋駕名04％6 ガ。屠。伽P π観。忽♂oπ（ゾ翫g1納．（4th edn）Edward Amold．
Jones， D．1949．7 加P名。η観磁 ガ。％（ゾEη8 1納． Cambridge University Press．
Jones， D．1960．ノ1η0％ 1加6げE％81ゴ∫乃P〃。π漉6∫． Heffer．
Kenyon， J． S．1950．んπ67z6碗Pηη観。磁。η． George Wahr．
Ladefoged， P．1969．ηz紹6∠レぞα∫σEκ1）67ゴ〃z6窺α1 Plzo麗 ゴcs． Oxford University Press．
Ladefoged， P．1982 A Coz6欝6勿P1乞。ηθ ガ。ε． Harcourt Brace Jovanovich．
Lewis， J． A．1972．．4 Coπo∫s6 P70πo瑚。加g 1万6 ガ。παηqプB7痂s〃απ4∠L吻67z6αηEηg1∫s肱Oxford University Press。
析谷 好弘．1976．「英語音声学」．こびあん書房，東京．
0 Connor， J． D．1988。 P加π6認。∫。 Penguin Books．
0 Connor， J． D．1990．β6飽プEηg1納P名。η朔6忽ガ。η． Cambridge University Press．
Ripman， W．1957．勘91納P勿勉∫6∫． J． M． Dent and Sons LTD．
Roach， P．1991． Eπg1納P加ηθ ガ6sαη4 Pぬ。π010gy． Cambridge University Press．
下田 弘之．1981． 「ai」「au」に関する一考察」 ．「英語の歴史と構造」宮部菊男教授還暦…記念論文集，
PP．395〜413，研究社．
Ward，1． C．1972．丁加P加勉∫6sσ翫g1ゴs乃． Heffer．
Wells， J． C．1982．且66翻sげE〃g1勅1：舳勉 剛〃吻π． Cambhdge University Press．
Wells， J． C．1990．五〇ηg初侃P名。η鰯6初加D纏。ηαη． Longman．
