Phonemic Transcription
This document describes an extension of the Trager-Smith phonemic alphabet for coding phonemic transcriptions of english (Old English, Middle English, Early Modern English, and Modern English).
The alphabet is derived from the ARPAbet (Shoup, 1980), extended for Middle English phonemes that are suitable for broad "phonemic" transcription. All phonemes are unigraphic or digraphic, chosen to avoid ambiguous interpretation when written without spaces between them. For example, all vowel phonemes are digraphs in which the first letter is "a", "e", "i", "o", or "u". For example the symbol 9 is used to represent The phoneme [ŋ] in singer is represented by "9" instead of "ng" to avoid confusion with separate articulation of the n and g, which are also phonemes. In addition to the phonemes of present day American English, the phonemes mentioned in Kökeritz, "A Guide to Chaucer's Pronunciation" which itself is similar to, but not identical with the current IPA standard alphabet.
The alphabet is designed to facilitate memorization and reading, but it is not proposed to be used by nonspecialists. Nor is it intended to be usable for "narrow" transcription. In fact, the vowels have been especially simplified.
The main design feature, aside from conservation of recognizable orthographic convention (based on the 26 letter latin alphabet), is that phonemes of a word can be written without a space between them without introducing ambiguities. The only ambiguity that I know in the original ARPABET is the [ng] phoneme which can be confused with the two phonemes [n g]. This phoneme symbol has therefore been changed to [9]. With this change, and respecting the principle with additions, any properly written string of phonemes has one and only one possible interpretation.
The rule for interpreting a written string of phonemes is simple to state. Reading from left to right, if a vowel (aeiou) is encountered, it is always the first half of a digraph. Reading from right to left, if an h is encountered then it is always the second half of a digraph, paired with the letter to its left. All other letters are unigraphs. Text written in this phonemic transcription can be translated into IPA symbols, a perl script that can be used to do this is here. A transcription can also be translated into the MBROLIGN text-to-speech system, and a perl script that can be used to do this is here.
Consonants
The following consonants are included in the alphabet. The IPA symbol, the symbol used in K, the SAMPA equivalent, the AB symbol, and examples of words from Pyles are shown.
| IPA | K | SAMPA* | AB | Notes |
|---|---|---|---|---|
| [b] | [b] | b | b | bib, ruby, rabble, ebb, tribe; |
| [p] | [p] | p | p | pup, stupid, aple, ripe; |
| [d] | [d] | d | d | dud, body, muddle, add, bride, seethed; |
| [t] | [t] | t | t | toot, booty, matter, butt, rate, hopped; |
| [g] | [g] | g | g | gag, lager, laggard, egg; |
| [k] | [k] | k | k | kit, naked, take, pick, mackerel, car, bacon, music; |
| [v] | [v] | v | v | valve, over; |
| [f] | [f] | f | f | |
| [ð] | [ð] | D | dh | then, either, eth, bethe; |
| [θ] | [θ] | T | th | thin, ether, froth; |
| [z] | [z] | z | z | zoos, fizzle, fuzz, ooze, visage, phase; |
| [s] | [s] | s | s | sis, pervasive, vise, passive, mass, cerial, acid, vice; |
| [ʒ] | [ʒ] | Z | zh | leisure, azuer, delusion, equation; genre, rouge; |
| [ʃ] | [ʃ] | S | sh | shush, marshal; |
| [h] | [h] | h | hh | ha, Mohawk; |
| [x] | [x] | x (k_h) | gh | ME only knight |
| [ç] | [ç] | C (k_h) | gh | ME only night |
| [ɣ],[ʝ] | [ç] | G (j) | yh | OE only fylgan 'follow' |
| [dʒ] | [dʒ] | dZ | jh | judge, major, gem, regiment, George, surgeon, region, budget; |
| [tʃ] | [tʃ] | tS | ch | church, lecher, butcher, itch; |
| [wh] | [hw] | wh (h w) | wh | [hw] in K, infrequently pronounced in Modern English, whether |
| [m] | [m] | m | m | mum, clamor, summer, time; |
| [n] | [n] | n | n | nun, honor, dine, inn, dinner; |
| [ŋ] | [ŋ] | N | 9 | sing, wringer, finger, sink; |
| [l] | [l] | l | l | lapel, felon, fellow, fell, hold; |
| [ɹ] [r] |
[r] | r | r | trilled [r] in ME, otherwise untrilled or allophonic, rear, baron, barren, err, bare; |
| [w] | [w] | w | w | won, which; |
| [j] | [j] | j | y | yet, bullion; |
Note that in all of the digraphs here, the second letter is always an h. When reading a transcription, an h that is not followed by an h belongs to the preceeding letter.
Alternatively, here are the above listed consonants in a matrix form.
| Manner of Articulation | Point of Articulation | ||||||
|---|---|---|---|---|---|---|---|
| Bilabial | Labio- dental |
Dental | Alveolar | Post- alveolar |
Palatal | Velar | |
| Stops | Voiceless Voiced |
p b |
t d | k g |
|||
| Affricates | Voiceless Voiced |
ch jh |
|||||
| Fricatives | Voiceless Voiced |
wh |
f v |
th dh |
s z | sh zh |
gh yh |
hh |
| Nasals | m | n | 9 | ||||
| Lateral | l | ||||||
| Retroflex | r | ||||||
| Semivowels | w | y | |||||
Pure Vowels
American English can be transcribed with 16 vowel phonemes, while British English requires an additional 3 (for softened [r] after vowels). Middle English seems to require even more vowel phonemes, roughly corresponding to variations heard in French and English. To the modern ear, many of these vowels are difficult to distinguish, and certainly today if not in Chaucer's day, might be considered allophonic with our modern vowels. Since our purpose is not to document subtle dialectal variations, but rather to give a broad transcription, it is necessary to simplify the ME vowels. These are listed below with their primary allophone. The first table lists "pure" vowels only, that is, non-diphthongs.
The following list of pure vowels is ordered approximately from front to back production. The long vowel indicator (:) is used in K as an integral feature of the phoneme. For example, in K, [i:] represents the IPA phoneme [i], while [i] represents [ɪ]. I think this is incorrect, the long duration indicator should not imply a variation of the sound, i.e. the formants. Therefore it will not appear in the IPA equivalents. With the exception of the shwa [ə], duration is not an feature of the phonemes. However, for translating back to the K notation, the long phoneme character will be used when the correspondence is clear.
| IPA | K | SAMPA* | AB** | Examples |
|---|---|---|---|---|
| [i] | [i:] | i | iy | evil, cede, meter, accretion, eel, lee, eat, sea; |
| [ɪ] | [i] | I | ih | it; mere, near, peer; illume, elude; aerial, area; topic, bucket, college; body, honey; |
| [e] | [e:] | ey | This phoneme is transcribed as an allophone of the diphthong [eɪ]. ape, basin, faint, gray; | |
| [ɛ] | [e] | E | eh | bet, threat; |
| [ɛ:] | eh | ? From K, there, Fr père | ||
| [æ] | [æ] | { | ae | at; |
| [a] | ae | ? From K, G Mann, Fr patte | ||
| [ə] [ɐ] |
[ə] | @ | ax | (schwa), this is an unstressed [ʌ]. alone; Cuba; bias; bursar; |
| [ʌ] | [ʌ] | V | ah | putt; |
| [ɑ] | [ɑ] | A | aa | father, art, stop; |
| [a:] | aa | ? From K, G Vater, Fr art | ||
| [ɒ] | [ɔ] | aa | ? BrE dot | |
| [ɔ] | [ɔ:] | O | ao | all, war, law, awe, cause, gone; |
| [o] | [o:] [o] |
o (@U) | oh | go, rode, road, toe, tow, owe, oh; |
| [ʊ] | [u] [U]2 |
U | uh | good, pull; |
| [u] | [u:] | u | uw | ooze, too, to, tomb, you, rude, rue, new; |
| [ü:] | uw | ? From K, Fr tu, G Tür; Like [ju]. |
Diphthongs and Other Vowels
In some circumstances, the consonants, r, m, n, l can form the nucleus of a syllable without a distinctive vowel. To transcribe these cases, we admit the following symbols.
| IPA | K | SAMPA | AB | Example |
|---|---|---|---|---|
| [ɚ] | [ə:] | 3 (r=) | er | bird |
| [m]* | [m]* | =m (@ m) | em | handsome |
| [n]* | [n]* | =n (@ n) | en | button |
| [l]* | [l]* | =l (@ l) | el | bottle |
If pure vowels are charactarized as sounds that are sustainable, a diphthong requires a varying sound. The diphthongs, when sustained, usually (in Modern English) sustain the second of two vowel phonemes. The first phoneme is pronounced transiently, almost imperceptible, making it difficult to identify it precisely. Diphthongs are always perceived as single syllables, unless spoken in an unnatural and exagerated way. For example going [gɔɪŋ] has two syllables while toy [tɔɪ] has the same sounds but only one syllable.
| IPA | K | SAMPA* | AB** | Phones | Example |
|---|---|---|---|---|---|
| [iə] | I@ (i @) | ia | iy + ax | BrE peer | |
| [ɪu] | [iu] | iu (I u) | iw | ih + uw | few |
| [ɛi] | [ei] | EI | ey | eh + iy | take, lake |
| [ɛə] | E@ (E @) | ea | eh + ax | BrE pair | |
| [ɛu] | [ɛu] | Eu (E u) | ew | eh + uw | |
| [ai] | AI | ay | aa + iy | ride, hie, my, style, stile, dye; | |
| [æi] | [æi] | ay | ae + iy | between lake and like | |
| [æu] | [ɑu] | aU | aw | ae + uw | how, house; |
| [ɔi] | [ɔi] | OI | oy | ao + iy | oil, boy; |
| [əI] | [əI]2 | @I (@ I) | ei | ax + ih | |
| [ʊə] | U@ (u @) | ua | uh + ax | BrE poor | |
| [əʊ] | [ɔu] [əU]2 |
@U | ow | ao + uw | oat |
| [ui] | ui (U i) | uy | uh + iy |
Other K phonemes
In the transcriptions of K, the word to is not transcribed, the vowel [o] being unspecified, but presumably this corresponds to the IPA phoneme. Since both long and short [o] are used in K, and the phoneme is the same with only the duration altered, and since our phonemic transcription does not reference phoneme duration, all [o:] in K will be changed to [o].
The phones [x] and [ç] are written differently in K, and indeed their sounds are different. However, they are not phonemically distinct (in fact, they are allophones of [h]). Both will be written [x]. The phonetic distinction is made by the vowel that preceeds the fricative, with the [ç] sound occurring after front vowels and [x] after back vowels.
To summarize the allophonic simplifications from K,
| From | To |
|---|---|
| [o:] | [o] |
| [ɛ:] | [e] |
| [ei] | [e:] |
| [æ] | [a] |
| [a:] and [ɔ] | [ɑ] |
| [ü:] | [u:] |
| [ç] | [x] |
Writing
Words are written without spaces between the phonemes. Syllable indicators and stress marks are optional.
There is a one-to-one correspondence between vowels (and ditphongs) and syllables. The association of consonants with the core vowel of a syllable is not exact, the goal is to make each syllable separately pronouncable. If there is a choice, consonants preceeding a vowel have a preference to stay with the vowel in the syllable. Stressed syllables will be preceeded by a stress mark (') for primary stress, or (,) for secondary stress, unstress syllables are preceeded by a dot (.), or by nothing for the first syllable of the word. Single syllable words do not require a separator or a stress mark, but the stress may be added to facilitate prose or poetry reading.
Pauses may be indicated with a vertical bar, one bar for a short pause and two bars for a longer pause. This is also helpful for poetry transcription. There is no good way to represent optional phonemes, so for the time being I include a note in parentheses that will require special interpretation.
Example Transcriptions
This is from K, translated into the AB alphabet.
whaen dhaet 'aap.rihl wihth ihz 'shuw.raxz 'soh.tax
dhax druwght axv maerch (hh)aeth 'peyr.saxd toh dhax 'roh.tax
aen(d) 'baa.dhaxd 'ehv.rih vayn ihn swihch lih'kuwr
axv whihch vehr'tuw ihn'jhehn.draxd ihz dhax fluwr |
whaen 'zeh.fih.ruhs eyk wihth ihz 'swey.tax brehth
ihn'spiy.raxd hhaeth ihn 'ehv.rih hhaalt aen(d) hhehth
dhax 'tehn.drax 'kraa.paxz | aen(d) dhax 'yuh9.gax 'suh.nax
hhaeth ihn dhax raem ihz 'hhael.vax kuwrs ih'ruh.nax |
aen(d) 'smaa.lax 'fuw.laxz 'maa.kaxn ,meh.loh'diy.ax
dhaet 'sley.paxn aal dhax niyght wihth 'ao.paxn 'iy.ax
sao 'prih.kaxth axm nae(:)'tuwr ihn ihr kuh'raa.jhaxz
dhaen 'laa(:)9.gaxn faalk toh gaon aan ,pihl.grih'maa.jhaxz
aen(d) 'paal.maxrz faar toh 'sey.kaxn 'strawn.jhax 'straa(:)n.daxz
toh 'fehr.nax 'hhael.waxz kuwdh ihn 'suhn.drih 'laa(:)n.daxz |
aen(d) 'spehs.yae.lih fraam 'ehv.rih 'shiy.raxz 'ehn.dax
axv 'eh9.gax.laa(:)nd toh 'kawn.taxr.brih dhay 'wehn.dax
dhax 'hhao.lih 'blihs.fuhl 'maer.tihr faar toh 'sey.kax
dhaet hhehm (hh)aeth 'hhaal.paxn whaen dhaet dhay wehr 'sey.kax ||
Notes
ae(:) may be aa
aa(:) may be ao
Here it is in IPA:
whæn ðæt 'ɑp.rɪl wɪθ ɪz 'ʃu.rəz 'so.tə
ðə druxt əv mærtʃ æθ 'peɪr.səd to ðə 'ro.tə
æn 'bɑ.ðəd 'ɛv.rɪ vaɪn ɪn swɪtʃ lɪ'kur
əv whɪtʃ vɛr'tu ɪn'dʒɛn.drəd ɪz ðə flur |
whæn 'zɛ.fɪ.rʊs eɪk wɪθ ɪz 'sweɪ.tə brɛθ
ɪn'spi.rəd hæθ ɪn 'ɛv.rɪ hɑlt æn hɛθ
ðə 'tɛn.drə 'krɑ.pəz | æn ðə 'jʊŋ.gə 'sʊ.nə
hæθ ɪn ðə ræm ɪz 'hæl.və kurs ɪ'rʊ.nə |
æn 'smɑ.lə 'fu.ləz 'mɑ.kən ,mɛ.lo'di.ə
ðæt 'sleɪ.pən ɑl ðə nixt wɪθ 'ɔ.pən 'i.ə
sɔ 'prɪ.kəθ əm næ'tur ɪn ɪr kʊ'rɑ.dʒəz
ðæn 'lɑŋ.gən fɑlk to gɔn ɑn ,pɪl.grɪ'mɑ.dʒəz
æn 'pɑl.mərz fɑr to 'seɪ.kən 'straʊn.dʒə 'strɑn.dəz
to 'fɛr.nə 'hæl.wəz kuð ɪn 'sʊn.drɪ 'lɑn.dəz |
æn 'spɛs.jæ.lɪ frɑm 'ɛv.rɪ 'ʃi.rəz 'ɛn.də
əv 'ɛŋ.gə.lɑnd to 'kaʊn.tər.brɪ ðaɪ 'wɛn.də
ðə 'hɔ.lɪ 'blɪs.fʊl 'mær.tɪr fɑr to 'seɪ.kə
ðæt hɛm æθ 'hɑl.pən whæn ðæt ðaɪ wɛr 'seɪ.kə ||
And here is the original K, after making the allophone substitutions (declared in the tables above) and removing the parenthetical notes.
hwan ðat 'ɑp.ril wiθ iz 'ʃu:.rəz 'so.tə
ðə dru:xt əv martʃ aθ 'pe:r.səd to ðə 'ro.tə
an 'bɑ.ðəd 'ev.ri væin in switʃ li'ku:r
əv hwitʃ ver'tu: in'dʒen.drəd iz ðə flu:r |
hwan 'ze.fi.rus e:k wiθ iz 'swe:.tə breθ
in'spi:.rəd haθ in 'ev.ri hɑlt an heθ
ðə 'ten.drə 'krɑ.pəz | an ðə 'juŋ.gə 'su.nə
haθ in ðə ram iz 'hal.və ku:rs i'ru.nə |
an 'smɑ.lə 'fu:.ləz 'mɑ.kən ,me.lo'di:.ə
ðat 'sle:.pən ɑl ðə ni:xt wiθ 'ɔ:.pən 'i:.ə
sɔ: 'pri.kəθ əm na'tu:r in ir ku'rɑ.dʒəz
ðan 'lɑŋ.gən fɑlk to gɔ:n ɑn ,pil.gri'mɑ.dʒəz
an 'pɑl.mərz fɑr to 'se:.kən 'strɑun.dʒə 'strɑn.dəz
to 'fer.nə 'hal.wəz ku:ð in 'sun.dri 'lɑn.dəz |
an 'spes.ja.li frɑm 'ev.ri 'ʃi:.rəz 'en.də
əv 'eŋ.gə.lɑnd to 'kɑun.tər.bri ðæi 'wen.də
ðə 'hɔ:.li 'blis.ful 'mar.tir fɑr to 'se:.kə
ðat hem aθ 'hɑl.pən hwan ðat ðæi wer 'se:.kə ||
For comparison, here is a transcription of Bessinger's recording, first in IPA:
'whɑn .ðæt 'ɑ.prɪl | .wɪθ .hɪz 'ʃəʊ.rəz 'səʊ.tə |
.ðə 'druxt .əv 'mɛrtʃ .hæð 'pɛr.səd .təʊ .ðə 'rəʊ.tə |
.ænd 'bɑð.əd 'ɛv.ri 'vaɪn .ɪn 'swɪtʃ .lɪ'kur
.əv 'whɪtʃ .vɚ'tu .ɛn'dʒɛn.dɚ.rɛd 'ɪz .ðə 'flur |
.whɑn .zɛ'fi.rəs 'eɪk | .wɪθ .hɪz 'sweɪ.tə 'brɛθ |
.ɪn'spi.rəd 'hæθ .ɪn 'ɛv.ri 'həʊlt .ʌnd 'hɛθ
.ðə 'tɛn.drə 'krɑ.pəz | .ænd .ðə 'jʊŋ.gə 'sʊ.nə
'hæθ .ɪn .ðə 'ræm .hɪz 'hæl.və 'kəʊrs .i'rʊ.nə |
.ænd 'smɑ.lə 'fu.ləs 'mæ.kən ,mɛ.ləʊ'di.jə |
.ðæt 'sleɪ.pɛn 'ɑl .ðə 'niɣt .wɪθ 'əʊ.pɛn 'i.jə |
.səʊ 'prɪ.kɛθ .hɛm .nɑ'tur .ɪn .hir .ku'rɑ.dʒɛz |
'ðæn 'lɑŋ.gʌn 'fəʊlk .tu 'gəʊn .ɑn 'pɪl.grɪ,mɑ.dʒʌz |
.ænd 'pɑl.mɚz .fəʊr .tu 'seɪ.kʌn 'stræn.dʒə 'strɑn.dʌz
.tu 'feɪr.nɛ 'hɑ.lwɛs 'kuð .ɪn 'sʊn.dri 'lɑn.dʌz |
.ænd 'spɛ.sjɑ.li | .frʌm 'ɛv.ri 'ʃi.rʌs 'ɛndə
.əv 'ɛŋ.gʌ.lɑnd .tu 'kɑn.tɚ.breɪ .ðaɪ 'wɛn.də |
.ðə 'həʊ.li 'blɪs.fʌl 'mær.tɚ .fəʊr .tu 'seɪ.kə |
.ðæt .hɛm .hæθ 'hɑl.pɛn 'whɑn .ðæt .ðeɪ .wɚ 'seɪ.kə |
And now in the K notation:
'hwɑn .ðat 'ɑ.pril | .wiθ .hiz 'ʃɔu.rəz 'sɔu.tə |
.ðə 'dru:xt .əv 'mertʃ .hað 'per.səd .tɔu .ðə 'rɔu.tə |
.and 'bɑð.əd 'ev.ri: 'væin .in 'switʃ .li'ku:r
.əv 'hwitʃ .və:'tu: .en'dʒen.də:.red 'iz .ðə 'flu:r |
.hwɑn .ze'fi:.rəs 'e:k | .wiθ .hiz 'swe:.tə 'breθ |
.in'spi:.rəd 'haθ .in 'ev.ri: 'hɔult .ʌnd 'heθ
.ðə 'ten.drə 'krɑ.pəz | .and .ðə 'juŋ.gə 'su.nə
'haθ .in .ðə 'ram .hiz 'hal.və 'kɔurs .i:'ru.nə |
.and 'smɑ.lə 'fu:.ləs 'ma.kən ,me.lɔu'di:.jə |
.ðat 'sle:.pen 'ɑl .ðə 'ni:xt .wiθ 'ɔu.pen 'i:.jə |
.sɔu 'pri.keθ .hem .nɑ'tu:r .in .hi:r .ku:'rɑ.dʒez |
'ðan 'lɑŋ.gʌn 'fɔulk .tu: 'gɔun .ɑn 'pil.gri,mɑ.dʒʌz |
.and 'pɑl.mə:z .fɔur .tu: 'se:.kʌn 'stran.dʒə 'strɑn.dʌz
.tu: 'fe:r.ne 'hɑ.lwes 'ku:ð .in 'sun.dri: 'lɑn.dʌz |
.and 'spe.sjɑ.li: | .frʌm 'ev.ri: 'ʃi:.rʌs 'endə
.əv 'eŋ.gʌ.lɑnd .tu: 'kɑn.tə:.bre: .ðæi 'wen.də |
.ðə 'hɔu.li: 'blis.fʌl 'mar.tə: .fɔur .tu: 'se:.kə |
.ðat .hem .haθ 'hɑl.pen 'hwɑn .ðat .ðe: .wə: 'se:.kə |
Prosody
For a natural, if not pleasing, computer reproduction of speech, it is necessary to annotate a few prosodic features. These include stress, syllable (vowel) duration, pause or junction, and pitch.
Our alphabet already makes use of stress indicators, /'/ for primary stress, /,/ for tertiary stress (in the sense of Trager-Smith), and /./ for no special stress. In addition to these, we add /^/ for secondary stress, which is defined to be an intentional elevation of tertiary stress, or intentional lowering of primary stress. We may assume that stress alters the pitch (within the 4 level scale) as well as duration and loudness. Stress marks apply to the following syllables. A space / / implies /./.
Vowels can be explicitly held for a longer duration by appending a /:/ character. This applies to the preceding vowel only.
Pauses will be indicated with a vertical bar, /|/, which indicates a normal pause. A longer pause can be specified with a double bar /||/. In addition, when it is necessary to indicate a broken pronunciation (minimal pause, also called a juncture), an underscore can be inserted /_/.
Finally, we assume 4 levels of pitch, from lowest to highest /1/, /2/, /3/, and /4/. These levels can be adjusted to a particular voice. Without specification, the pitch level is assumed to be /1/. The pitch indicator applies to all syllables following it, until another pitch indicator is given. In addition, the indicators /-/ and /+/ will be used to indicate that the syllable or vowel immediately preceding is to have a falling (/-/) or rising (/+/) pitch.
For a complete specification of pitch, We will experiment with a melodic annotation that uses the frequency of the pitch of each vowel. Since the style of vowel pronunciation may have a variable tone, we allow up to three notes to be supplied with each vowel. Consonants are pronounced by a smooth transition between the surrounding tones. Melodic data is provided by specifying frequencies (in Hertz) for each syllable (vowel) between square brackets. Specifications for vowels are separated by semi-colons. Multiple notes per vowel are separated by commas.
We will allow melodic data to appear anywhere in a transcription, always interpreted as applying to the most recent preceeding vowels. A complete specification will have one or more notes given for each vowel in the transcription. Given the one-to-one correspondence of notes to vowels, we can also permit stress indicators (') and (,) as well as the long duration indicator (:) to appear in the melodic transcription.
Hybrid Transcription
Here is a hybrid based on the phonemic transcription of K, translated into this alphabet, and prosody imitating Bessinger. Stresses for multisyllabic words are the same as they appear in K, and pauses in K also appear here. Stress have only been added to one syllable words with metrical stress, and pauses added, where it seems natural after Bessinger.
'whaen dhaet 'aap.rihl | wihth ihz 'shuw.raxz 'soh.tax |
dhax 'druwght axv 'maerch (hh)aeth 'peyr.saxd ,toh dhax 'roh.tax |
aen(d) 'baa.dhaxd 'ehv.rih 'vayn ihn 'swihch lih'kuwr
axv 'whihch vehr'tuw ihn'jhehn.draxd ihz dhax 'fluwr ||
whaen 'zeh.fih.ruhs 'eyk | wihth ihz 'swey.tax 'brehth |
ihn'spiy.raxd 'hhaeth ihn 'ehv.rih 'hhaalt aen(d) 'hhehth
dhax 'tehn.drax 'kraa.paxz | aen(d) dhax 'yuh9.gax 'suh.nax
hhaeth ihn dhax 'raem ihz 'hhael.vax 'kuwrs ih'ruh.nax ||
aen(d) 'smaa.lax 'fuw.laxz 'maa.kaxn ,meh.loh'diy.ax |
dhaet 'sley.paxn 'aal dhax 'niyght wihth 'ao.paxn 'iy.ax |
sao 'prih.kaxth axm nae'tuwr ihn ihr kuh'raa.jhaxz |
dhaen 'laa9.gaxn 'faalk toh 'gaon aan ,pihl.grih'maa.jhaxz |
aen(d) 'paal.maxrz faar toh 'sey.kaxn 'strawn.jhax 'straan.daxz
toh 'fehr.nax 'hhael.waxz | 'kuwdh ihn 'suhn.drih 'laan.daxz ||
aen(d) 'spehs.yae.lih | fraam 'ehv.rih 'shiy.raxz 'ehn.dax
axv 'eh9.gax.laand toh 'kawn.taxr.brih dhay 'wehn.dax |
dhax 'hhao.lih 'blihs.fuhl 'maer.tihr faar toh 'sey.kax |
dhaet hhehm (hh)aeth 'hhaal.paxn | 'whaen dhaet dhay wehr 'sey.kax ||
bih'fehl dhaet ihn dhaet 'seh.zuhn aan ae 'day |
ihn 'suh.dhaxrk aet dhax 'tae.baerd aez ih 'lay |
'reh.dih toh 'wehn.daxn aan mih ,pihl.grih'maa.jhax
toh 'kawn.taxr.brih | wihth 'fuhl dax'vuwt kuh'raa.jhax |
aet 'niyght waez 'kuhm ihn.toh dhaet ,aas.taxl'riy.ax |
wehl 'niyn aen(d) 'twehn.tih ihn ae ,kuhm.pay'niy.ax
axf 'suhn.drih faalk | bih ,ae.vaxn'tuwr ih'faa.lax
ihn 'feh.law.shihp | aen(d) 'pihl.grihmz 'wehr dhay 'aa.lax
dhaet 'toh.waerd 'kawn.taxr.beh.rih 'wohl.daxn 'riy.dax ||
dhax 'chaam.braxz aen(d) dhax 'staab.laxz 'weh.raxn 'wiy.dax |
aen(d) wehl weh 'weh.raxn 'eh.zaxd ae.tax 'behs.tax ||
aen(d) 'shaar.tlih | whaen dhax 'suh.nax waes toh 'rehs.tax |
soh hhaed ih 'spao.kaxn wihth axm 'ehv.rihch'aon |
dhaet 'iy waez axv ihr 'feh.law.shihp ae'naon |
aen(d) 'maa.dax 'faar.waerd 'ehr.lih faar toh 'riy.zax
toh 'taak uwr 'way | 'dhehr aez ih 'yuw deh'viy.zax ||
buht 'nae.dhax'lehs | whiyl ih (hh)aev 'tiym aen(d) 'spaa.sax |
ehr dhaet ih 'fehr.dhaxr ihn dhihs 'taa.lax 'paa.sax |
meh 'thih9.kehth iht ae'kaar.dawnt toh reh'zuwn
toh 'teh.lax yuw 'aal dhax kaan,dih.sih'uwn
axv 'ehch axf hhehm | saa 'aez iht 'sey.maxd 'mey |
aen(d) 'whihch dhay 'weh.raxn | aend axv 'whaet dax'grey |
aend 'eyk ihn 'whaet ae'ray dhaet dhay wehr 'ih.nax |
aend aet ae 'kniyght dhaen wuhl ih 'fihrst bih'gih.nax ||
Here is some approximate pitches (in Hz), for each of the vowels in the first stanza, separated by semicolons. Multiple tones per vowel are separated by commas. This has been used with the phonetic transcription above to synthesize the first stanza using MBROLA, the result is a computer sounding voice that imitates Bessinger's intonation and pronunciation suggested by K.
[210,250;220;170;75;135;165;180,215;210;155;155;]
[159;230;170;190,80;135;130;108;120;120;90;100;]
[220;190;180;180;180;170,180;170;170;165;170;]
[140;150;170;160;150,140;145;140;132;145;120,90,114;]
[105;150;110;110;168,155;126;137;120,160,150;140;145;]
[134;169;177;173;170;190;183;130,190;174;147,190,190;]
[184;150,194;163,186;190,135;140,130;94,148;155,140;139,161;157;165;148;]
[178;161;149;130,173,143,123,120;152;150;138;127,110;131,100;105;109,120;]
[108;117,144;138;133;117;120;116;117;115;109;93,81;]
[94;108;99;115,145;150;150,105;108;142;96;111;90;]
[95,85;113;91;87;87;121,89;90;90;90;80;85;]
[111,182;170;156;174,152;138;146;142;155,117;104;104;111;]
[128,146;194,180,92;84;113;118;130;113;117,142,137;134;129;127;]
[131;126;116;112,94;82;110,96;95;108;94;88;77;]
[89;142;115;98;122,140;154,178;171,168;166;164;164;150;]
[124,151;149;146;129;132;142,166,138;112,96;97;98;100;112;]
[125;110,132,134;127;121;119;110;83;91;92;83;81;]
[91;92;95;115,102;75;81;82;92,75;80;124,90;75;]
The text-to-speech from by MBROLIGN using the K phonemes and the Bessinger intonation can be heard here.
Shakespeare
Sonnet XVIII
ʃæl əɪ kəm'pɛr ði: tʊ ə 'sʌ.mərz 'dɛ: |
ðəʊ æ:rt 'mo:r 'lʌv.lɪ ən 'mo:r 'tɛm.pə,rɛ:t ||
'rʌf 'wəɪnz du 'ʃɛ:k ðə 'dær.lən 'bʌdz əv 'mɛ: |
ən 'sʌ.mərz 'lɛ:s əθ 'o:l tu: 'ʃɔ:rt ə 'dɛ:t ||
'sʌm.təɪm 'tu: 'hɑt ðɪ 'əɪ əv 'hɛ.vn 'ʃəɪnz |
ən 'o:.fn ɪz ɪz 'go:ld kəm'plɛk.ʃn 'dɪmd |
ən 'ɛv.rɪ 'fɛ:r frəm 'fɛ:r sʌm'təɪm dɪ'kləɪnz |
bɪ 'tʃæ:ns ər 'nɛ:.tərz 'tʃɛ:n.ʒn 'ko:rs ʌn'trɪmd ||
bət 'ðəɪ ɪ'tæ:r.nəl 'sʌ.mər 'ʃæl 'nɑt 'fɛd |
no:r 'luz pə'zɛ.ʃn əv ðæt 'fɛ:r ðəʊ 'o:st |
no:r ʃæl 'dɛ:θ 'bræg ðəʊ 'wɑn.dərst ɪn ɪz 'ʃɛ:d |
hwɛn ɪn ɪ'tæ:r.nəl 'ləɪnz tə 'təɪm ðəʊ 'gro:st ||
so: 'lɑŋ əz 'mɛn kən 'brɛ:ð ər 'əɪz kən 'si: |
'so: 'lɑŋ 'lɪvz 'ðɪs ən 'ðɪs gɪvz 'ləɪf tʊ 'ði: ||
Sonnet XXX
'hwɛn tu ðə 'sɛ.ʃnz əv 'swi:t 'səɪ.lənt 'θɔ:t
əɪ 'sʌ.mən 'ʌp rɪ'mɛm.brəns əv 'θɪŋz 'pæ:st |
əɪ 'səɪ ðə 'læk əv 'mɛnj ə 'θɪŋ əɪ 'sɔ:t
ən wɪð 'o:ld 'wo:z 'nju: 'wɛ:l mɪ 'di:r 'təɪmz 'wɛ:st |
'ðɛn kæn əɪ 'drəʊn ən 'əɪ ʌn'juzd tə 'flo: |
fɜ: 'prɛ.ʃəs 'frɛnz 'hɪd ɪn 'dɛθs 'dɛ:t.lɪs 'nəɪt |
ən 'wi:p ə'frɛʃ 'lʌvz 'lɑŋ sɪns 'kæn.səld 'wo: |
ən 'mo:n ðɪk'spɛns əv 'mɛnj ə 'væ.nɪʃt 'səɪt ||
'ðɛn kæn əɪ 'gri:v ət 'gri:.vən.sɪz fo:r'go:n |
ən 'hɛ.vɪ.lɪ frəm 'wo: tʊ 'wo: 'tɛl 'o:r
ðə 'sæd ə'kəʊnt əv 'fo:r.bɪ'mo:.nɪd 'mo:n |
hwɪtʃ əɪ 'nju 'pɛ: əz ɪf nɑt 'pɛ:d bɪ'fo:r ||
bət ɪf ðə 'hwəɪl əɪ θɪŋk ɑn 'ði: | 'dir 'frɛnd |
'ɔ:l 'lɔ:.sɪz æ:r rɪ'sto:rd ən 'sɑ.roz 'ɛnd ||
Sonnet CXVI
'lɛt mi: 'nɑt tə ðə 'mæ.rɪdʒ əv 'tru: 'məɪnz
əd'mɪt ɪm'pɛ.dɪ.mənts || 'lʌv ɪz 'nɑt 'lʌv
hwɪtʃ 'ɔ:l.tərz hwɛn ɪt ,ɔ:l.tə'rɛ:.ʃn 'fəɪnz |
ɔr 'bɛnz wɪð ðə rɪ'mu:.vər tʊ rɪ'mu:v ||
'o: | 'no: | ɪt ɪz ən 'ɛ.vər 'fɪk.sɪd 'mæ:rk
ðət 'lʊks ɑn 'tɛm.pɪsts ənd ɪz 'nɛ.vər 'ʃɛ:.kn ||
ɪt ɪz ə 'stæ:r tʊ 'ɛv.rɪ 'wɑn.drɪn 'bæ:rk |
hu:z 'wɜ:rθs 'ʌn'no:n | ɔ:l'ðo: ɪz 'həɪθ bɪ 'tɛ:.kn ||
'lʌvz nɑt 'təɪmz 'fu:l | ðo: 'ro:zɪ 'lɪps ən 'tʃi:ks
wɪ'ðɪn ɪz 'bɛn.dn 'sɪ.klz 'kʌm.pəs 'kʌm ||
'lʌv 'ɔ:l.tərz 'nɑt wɪð ɪz 'bri:f 'o:rz ən 'wi:ks |
bət 'bɛ:rz ɪt 'əʊt 'i:n tə ðɪ 'ɛdʒ əv 'du:m ||
ɪf 'ðɪs bi 'ɛ.rər | ənd ə'pɑn mi 'pru:vd |
əɪ 'nɛ.vər 'rɪt | nɔ:r 'no: 'mæn 'ɛ.vər 'lʌvd ||
About this document
This document was written and copyright (C) 2006 by Larry Smith. The content is an original exploration of the phonemes of english from a perspective of a synthesis of their historical forms.