Update: 2020-05-09 06:43 PM -0400

TIL

Vowels and Consonants
of BEPS languages

MC-acoustics.htm

by U Kyaw Tun (UKT) (M.S., I.P.S.T., USA) and staff of Tun Institute of Learning (TIL) . Not for sale. No copyright. Free for everyone. Prepared for students and staff of TIL Computing and Language Center, Yangon, MYANMAR : http://www.tuninst.net , www.romabama.blogspot.com

Addenda to edited version of A Practical Sanskrit Distionary, by A. A. Macdonell, 1893,
http://www.sanskrit-lexicon.uni-koeln.de/scans/MDScan/index.php?sfx=jpg; 1929.
Nataraj ed. (reprint of 1914ed.), 1st in 2006, 2012.

index.htm | Top
MC-indx.htm

Contents of this page

UKT 161030: Read also Vowels in Human Voice, Phonetics and Phonology:
- HV-indx.htm > vow.htm (link chk 161030)

• Abugida-Akshara system of writing : different from Alphabet-Letter system
• Modification of vowel sound or syllable : {sa.ka:þän-su.}
• Similar vowels and dissimilar ones : {þa.wûN} & {a.þa.wûN}
• The problem of "Open O" /ɔ/ : how script might have changed
without much change in pronunciation. The solution is probably in
the {þa.wûN} 'ill-matched' pairs in Bur-Myan cf. Mon-Myan
• Voice quality

UKT notes :
• Intrinsic vowel and its pronunciation

Contents of this page

Abugida-Akshara system of writing :
different from Alphabet-Letter system

-- UKT 130828, ... , 161029, 200430

Editing a dictionary on Indic & Myanmar languages is a mess until one keep in mind that the medium of presentation to the public is Eng-Lat (English speech written in modified Latin). Other west European languages also write in other forms of modified Latin, such as French-Latin, German-Latin, Italian-Latin, and Spannish-Latin. However the basic Latin alphabet is the same. The basic unit is the Letter which is mute. Make sure to differentiate English from French, German and Spanish: write Eng-Lat (English-Latin), French-Latin, German-Latin and Spanish-Latin. The speech of Rome is written in Latin, and thus we should write Roman-Latin. We should also say "Latinization" in stead of "Romanization".

Eng-Lat writes in Alphabet-Letter writing system, where the fundamental unit, the Letter is mute. For instance, the Letter <t> /t/ cannot be pronounced. However if you couple it with a vowel, say <a> /a/, the combination <ta> /ta/ is pronounceable.

Also remember that among the European languages which write in basic Latin script, Eng-Lat is notorious for being non-phonetic: there is no correspondence between speech and script. Pronounce an English word as it is spelled and you are in trouble.

Other languages in BEPS - Bur-Myan, Pal-Myan, & Skt-Dev, use the Abugida-Akshara writing system. The fundamental unit, the Akshara is pronounceable. We say it is a syllable {sa.ka:þûn-su.}. The aim of these writing systems are to have a one-to-one correspondence between speech and script. Moreover this one-to-one correspondence must never change. Because of this unchanging or ever-lasting nature, both the basic element in speech and script are known as Akshara {ak~hka.ra}. Thus Abugida should be specified as Abugida-Akshara.

However, it should be remembered that though two languages use the same basic script, two words spelled the same are pronounced differently. Thus <cha> in French as {sha.} /ʃæ/ whereas in English it is {hkya.} /ʧæ/. The two are pronounced differently because of the differing phonologies. Since , both {sha.} /ʃæ/ and {hkya.} /ʧæ/, are base-consonants and are stable under the Virama {a.þût}, Romabama has to use special glyphs such as {sha.}/ {sh}, and {cha.}/ {ch}. Thus Romabama can transcribe the English word <church> as {chaach}.

When I use Mon-Myan to test the effectiveness of Romabama, it fails miserably because Bur-Myan and Mon-Myan, in spite of the same akshara uses different phonologies just as in English and French.

Mon-Myan is of interest because, there can be two types of intrinsic-vowels for consonants. See more of intrinsic vowels in
- MC-in-vow.htm (link chk 200430)

Velar {ka.} row #1 - bk-cndl-Mon-row1<)) (all links checked 200430)
Palatal row #2 - bk-cndl-Mon-row2<))
Mon-Myan pronounces as Palatal-affricates {ca.} / {cha.},
whereas Bur-Mon as Palatal-stops {sa.} / {hsa.}
Dental {ta.} row #4 - bk-cndl-Mon-row4<))
Labial {pa.} row #5 - bk-cndl-Mon-row5<))

Bangla-Bengali is of interest because it has split vowels just as in Bur-Myan and Mon-Myan. See more of split vowels in - MC-syllab.htm (link chk 150924)

Since the aim of abugidas is a one-to-one correspondence between speech and script, using ordinary Eng-Lat for transcription usually brought in misunderstandings and confusion. Because of this transliteration is mainly used instead of transcription. However when readers pronounce according to transliteration many native words became misunderstood or have their meanings changed. So the first step is to write English in modified Latin script such as IPA to bring on one-to-one correspondence. TIL uses Eng-IPA-Latin (or Eng-IPA) in BEPS work to compare Sanskrit words to Pali.

The two writing systems, the Abugida (for Burmese, Pali, & Sanskrit) and the Alphabet (for English, Georgian & Roman), are entirely different. The basic unit of script in Abugida is the syllable, where the consonant and vowel characters, are specified as Aksharas. The basic consonants have an inherent-vowel which is never specified. At present this inherent vowel is taken to have the the pronunciation /ə/. For transliteration and transcription to common English the intrinsic vowel is marked as <a> /a/. However, it never given as a glyph in Myanmar and Devanagari. Only a faint circle is given in Devanagari, and only a short horizontal line in Myanmar, e.g. the intrinsic vowel of {pa.} is not given, but "marks" to indicate the changes of the sound are given:

{a.} अ , {a} आ
{pa.} प , {pa} पा , {pi.} पि , {pi} पी , {pu.} पु , {pu} पू .

For representing Skt-Dev, see Unicode 4.0, ch09 - Unicode<Ô> (link chk 150908, not working in 190803)

These marks are known as vowel-signs, which I sometimes shorten to vow-signs. The vowel-signs are used to modify consonants. When we need vowels as independent characters, we use vowel-aksharas (or vowel-letters).

The basic unit of script in an Alphabet is the Letter without an inherent vowel. Thus the consonant {pa.} is pronounceable, whereas p is not.

Since you will come across more of the difference between Abugida and Alphabet in
Comparison of Skt-Dev, Eng-IPALatin, and Bur-Myan vowels - MC-BEPS-vow.htm (link chk 150820)
I will conclude this section with a modern sample of writing Myanmar script written on a leaf of Talipot palm, Corypha umbraculifera. The characters are scratched on the leaf using a steel stylus. The marks are made more visible by rubbing in lamp-black into the scratches. The lamp-black is in the form of a paste in crude petroleum oil. The leaf becomes yellow because of the oil, but over time turns brown.

Writing on palm leaves with a stylus is still practiced in Myanmarpré by astronomers-astrologers for each person giving the exact time of date, day and time of birth based on the Myanmar luni-solar calendar. Since, the date depends on a particular luni-solar calendar, which has seen changes during the long history of the country, the positions of the planets, and the asterisms are calculated and recorded. Do not forget that the Western calendar has been changed within our living memory, and changes to the dates of birth of historical persons, and historical events have to be specified in BC or AD. Similarly, the time-keeping devises have been changed and so when recording the "hour, minute, and second" of birth, the time-keeping system must be specified. The Western historians are only now beginning to realize the utility of such a system for keeping track of historical events which could be checked by modern day astronomy.

Contents of this page

Modification of vowel sound or syllable

Modifications can occur to the vowel sound or syllable, both at the beginning of the sound is the onset consonant, and at the end or the coda consonant. This definition is important in the transcription of Burmese-Myanmar into Romabama {ro:ma.ba.ma}.

How the modification occurs depend on the vowel itself, which is pronounced differently in different languages. For example, a basic vowel such as /i/ is pronounced similarly but differently in different languages because of the muscles being used. This results in what I am calling the Curse of the Babel.

In Romabama we are faced with the problem to represent the 3 pitch-registers of Bur-Myan with two tones (short vowels of duration 1 eye-blink and long vowels of duration 2 blk) of Eng-Latin, and we run smack into what I had termed the Two-three tone problem. However inclusion of Mon-Myan has solved this problem, e.g.:

{a:.} (1/2 blk), {a.} (1 blk), {a} (2 blk), {a:} (2 blk with emphasis)

Mon-Myan has {a:.} (1/2 blk), {a.} (1 blk), {a} (2 blk),
Eng-Lat has {a.} (1 blk), {a} (2 blk), and
Bur-Myan has {a.} (1 blk), {a} (2 blk), {a:} (2 blk with emphasis).
Bur-Myan also has {aa.} which is equiv to Mon-Myan {a:.} (1/2 blk).

UKT 150925 : CAUTION: What English ESL has termed "short vowels" and "long vowels" are misnomers, as they have nothing to do with time-duration. See also in this folder Short Vowels /æ/ - ESL-aeiou<Ô> (link chk 150920)
and listen to - ESL-aeiou<)) (link chk 150917)
and identify:
<a> /æ/ in <cat> /kæt/ ------ {kakt}
<e> /e/ in <bed> /bed/ ------- {bèd} (cf. {hkít} 'era')
<i> /ɪ/ in <pig> /pɪg/ ---------- {pig}
<o> /ɒ/ in <dog> /dɒg/ (US) /dɑːg/ - {daug}
<u> /ʌ/ in <bus> /bʌs/ -------- {bûtS}
A more complete list will be found in
• Section 1. English pronunciation guide - EPG-indx.htm (link chk 150920)
and go to ¤ Vowels - EPG-vow.htm (link chk 150920)
My usual approach is to study Short Vowels by taking Vowel-Coda , VC (or VÇ), together. Never forget that the Canonical structure of Eng-Lat syllable is CVC, and that of Bur-Myan is CVÇ

Before proceeding further, it should be noted that the line between vowels and consonants, in terms of sonority, cannot be clearly drawn; a continuum exists between the two extremes. Many terms have been created, such as free & checked vowels, tense (long) and lax (short) vowels, obstruents and the sonorants, spirants. Now just go by the sonority scale , and my modified table of IPA pulmonic consonants. And instead of trying to reconcile the different usages, I have confined myself to two definitions only.
1. Inherent vowel : Each basic consonant-Akshara, such as /ta/ has an inherent vowel which is usually equated to short English a . In Bur-Myan, Pali-Myan, and Skt-Myan, it is {a.}. But in Mon-Myan, we find {é} in addition to {a.}. In Eng-Lat, the basic unit is the Letter, and it is mute. Thus instead of /ta/, the letter is only /t/.
2. Nuclear vowel: Each word of Bur-Myan and Skt-Dev has the Canonical CVÇ, where C is onset and can have the values 0, 1, 2, 3. V is the nuclear vowel of the syllable. Ç is coda, and can be 0, 1, 2, 3. In indigenous Bur-Myan Ç can be 0 or 1 only. Pal-Myan has only vertical and horizontal conjuncts which do not show Ç explicitly. Eng-Lat treats C and Ç the same: its canonical structure is CVC.

Now, what are Bur-Myan {sa.}/ {c} (r2c1) and {þa.}/ {þ} (r6c5) (usually transcribed as {tha.}/{þa.}? And, Skt-Dev स (U0938) pronounced as /s/ which I have found to be equal to Bur-Myan {þa.} /θ/. Are they sibilants or thibilant, or something else? Finally, what are श ś [ɕ] /ʃ/ and ष ṣ [ʂ] /s/? . These two have now been incorporated into BEPS as {sha.}/ {sh} and {Sa.}/ {S}.

Whenever we come across Eng-Latin <p, t, k>, we must be careful about the pronunciation of /p/, /t/, and /k/. English pronunciation of the individual grapheme is of two kinds which are called allophones. For example, the letter <p> with the pronunciation /p/ (broad transcription) has two allophones represented in narrow transcription: [p] in <spin> corresponding to {pa.} sound, and [pʰ] in <pin> corresponding to {hpa.} sound.

The English allophones [p t k] correspond to ,, {pa. ta. ka.}, and [pʰ tʰ kʰ] to ,, {hpa. hta. hka.} . Allophones ,, {pa. ta. ka.} are known to TIL as tenuis consonants and are in column c1 of akshara matrix, and ,, {hpa. hta. hka.} as voiceless consonants and are in c2 . The voiced consonants in c3 are ,, {ba. da. ga.}, and deep-H or glottal consonants in c4 are ,, {Ba. Da. Ga.}. Do not use "aspirate" for c2 and c4. To us they are phonemes in their own right and do not have aspirated sounds. We usually list our phonemes starting from interior of the mouth and proceed to outer parts. Our order is velar-to-labial, but the Westerners list theirs in order labial-to-velar. See IPA consonantal table. However, because I am using Eng-Lat as a means of communication, I have to follow their order:

,, r5c3 r4c3 r1c3 ,, r5c4 r4c4 r1c4
,, r5c1 r4c1 r1c1 ,, r5c2 r4c2 r1c2

The above 12 are the aksharas and letters of plosive-stop consonants we need most for inter-transcription between Bur-Myan and Eng-Latin. If I were to arrange the order of above:

,,, r1c1 r1c2 r1c3 r1c4
,,, r4c1 c4c2 r4c3 r4c4
,,, r5c1 r5c2 r5c3 r5c4

Eng-Latin has only two nasals, <m> /m/ & <n> /n/, and the nasal /ŋ/ in words like <king> /kɪŋ/. If we are to include these 3 as well, we have the following. However, it will not do in Bur-Myan (for BEPS), because it is in the pronunciation of "killed" consonants that we have trouble connecting to Eng-Latin :

,,,, ---- r1c1 r1c2 r1c3 r1c4 r1c5 ------- ,,,,
,,,, ---- r4c1 c4c2 r4c3 r4c4 r4c5 ------ ,,,,
,,,, ---- r5c1 r5c2 r5c3 r5c4 r5c5 ------ ,,,,
Bur-Myan (for pure Burmese) is simpler in the killed consonants since only, , ; ,

In many English words, <g> is not pronounced, in which case <ng> (perhaps best represented as <n^g>) behaves exactly like Bur-Myan {nga.}. Remember, <ng> is a digraph, and the "g" is generally not pronounced: <singer> /sɪŋ.əʳ/ (US) /sɪŋ.ɚ / vs. <finger> /fɪŋ.g|əʳ/ (US) /fɪŋ.g|ɚ /. In these two words, <ng> is pronounced differently.

To render Eng-Lat <ing> into Romabama - keeping in mind that <g> is silent - we can look the following illustration:

Eng-Lat: <kin> /kɪn/ (DJPD16-300), vs., <king> /kɪŋ/ (DJPD16-300)
Bur-Myan: {kín:} ---------------------------- {kaín:}

Now compare them to {kûn} and {kän}. There is no confusion between {kïn-} and {kûn-} or {kän}.

Remember the role of <n> in three different words. <n> means differently: it must be taken together with the preceding vowel. Even if the diacritical marks are lost, there is no confusion.

{kïn-} --> {kin}
{kûn} --> {kun}
{kän} --> {kan}

This is the aim of Romabama: even if the diacritical marks are lost, Romabama transcription must still give a one-to-one correspondence to Bur-Myan spelling.

IPA arranges the oral plosives in the order p --> k, whereas in the aksharas including Devanagari, the usual way is k --> p. I am following the akshara way. Below is an expansion of IPA consonants:

Contents of this page

Similar vowels and dissimilar ones
{þa.wûN} & {a.þa.wûN}

UKT 150925: Perhaps it's time for me to eat crow, and make a major correction
to Romabama in connection with A'thawun vowels. I still need to study more of
both Bur-Myan and Mon-Myan.

- UKT 130828, ...150926

I have state above that based on this idea of vowel duration, measured in time-duration you take to blink your eye, we can find which vowels can form {þa.wuN} 'similar pairs' or beautifully matched pair, and, which when paired form {a.þa.wuN} 'dissimilar' or ill-matched pair.

Vowels at the corners of the vowel quadrilateral, /a/, /i/, /u/, are long vowels of duration 2 blks. They have short counterparts with duration 1 blk which are nearer to the center. They are the ones which form {þa.wuN} 'similar or matched pairs':

{a.}, {a},
{i.}, {i},
{u.}, {u}.

In the vowel quadrilateral, there are four back vowels as a group, /u/, /o/, /ɔ/ & /ɑ/. The problem of relating Bur-Myan to Eng-Lat becomes worse when there is a mix up of /ɔ/ & /ɑ/.

Remember our interest is in mid-vowels: in Myanmar, for the front we have {è} and {e}, and for the back {u}, {o}, {AU}/{AW}, & {au}. But, English has only one for the front, <e>, and three, <u>, <o>, <au>. Note: <au> is a digraph and is pronounced as a monophthong in Bur-Myan.

We now look for a possible solution in English pronouncing dictionary - DJPD16-indx.htm (link chk 150926),
and proceed to Pronouncing the letters: vowels
- let-a.htm , let-e.htm , let-i.htm , let-o.htm & let-u.htm (link chk 150814).

UKT 150926:
Since my present transcriptions {è}/{ei} - {e}, and, {AU}/{AW} - {au} , have called for <ei>, <e> and <au>, I will look into:

letters EI
p173. There are several pronunciation possibilities for the vowel diagraph [ei].
vowel-sound /eɪ/ or /aɪ/
<eight> /eɪt/
<height> /haɪt/

letter E
p168. The vowel letter [e] has two main strong pronunciations linked to spelling: a 'short' pronunciation /e/ and a 'long' pronunciation /iː/. However, the situation is not clear cut and other pronunciations are available.
vowel-sound: /e/ -- short pronunciation
<bed> /bed/
<bedding> /ˈbed.ɪŋ/

letters AU, AW
p41. The vowel letter combinations [au] and [aw] are similar in that their most common pronunciation is /ɔː (US) ɑː/ , e.g.:
vowel-sound /ɔː/ (US) /ɑː/
<sauce> /sɔːs/ (US) /sɑːs/
<saw> /sɔː/ (US) /sɑː/

A possible help to solve our problem is in tense/lax dimension of the vowels in English. See my notes below.

Contents of this page

The problem of "Open O" /ɔ/ :

how script might have changed without much change in pronunciation

- UKT 150821, 150925

To get a fuller understanding on /ɔ/ (U0254) known as "Open-O", which could be {AU}/{AW}, I need to study Mon-Myan phonology by repeated listening to the sounds of phones in isolation, in continuous speech, and in songs. Let's listen to the vowels first paying attention to the A'thawun {a.þa.wûN} 'ill-matched' pairs. In Mon-Myan, the three {þa.wûN}-pairs are:
- {É è}, - {AW é}, - {än a:.} - Mon-v2<)) (link chk 150924)

You will notice that Bur-Myan needs a new phone, {é}, which is usually mis-identified as {au}. We can now identify the contrastive front-back mid-vowels as {e}~ {o}, and {è}~ {é}.

UKT 150925: The bookmarks for contrastive vowel-pair, important for TIL editors, have been set for:
- {è} as <ei3>, and {é} <ou2>.
See English contrastive vowels, by P. Ladefoged in TIL SD-Library
- Ladefoged-Eng-Contra-vow<Ô> (link chk 150926)
to know what is meant by "contrastive vowels". Ladefoged gives only h_d (onset-h & coda-d), such as
[he^ɪd] <hayed>, and [ha^ʊ] <how> . His examples are not applicable to our problem.

Both Mon speech and Burmese speech uses the same basic Myanmar script, similar to French and English speeches using the basic Latin script. Myanmar script is unique in being the only language based on circularly rounded circles, except perhaps Georgian script : თ [Georgian Alphabetic-Letter Tan] & ი [Georgian Alphabetic-Letter In]. If we are to take that the Georgian vowel In ი is the equivalent of Bur-Myan vowel-sign {loän:kri:tïn} which has its position changed, , the correspondence is complete between the Georgian word თი and Bur-Myan {ti.}, both pronouncing /ti/. This is perhaps how an Abugida has been changed into an Alphabet.

Mon-Myan and Bur-Myan are completely different languages. They even belong to different language groups. Mon belongs to Austro-Asiatic language group and Burmese to the Tibeto-Burman. Yet, because speakers of both languages are by religion, Theravada, they think the same way. They eat the same diet, and treat their women folk the same with what might amount to reverence. They place their mothers in place of honour, and treat their wives almost the same as themselves. To them the land they live on is their Mother-land. To them the country they live in is A'Mi Myanmarpré {a.mi.mrun-ma-præÑ}. - not their "Father Land".

In a way both Bur-Myan & Mon-Myan are maternalistic societies -- a trait which they probably acquired from the Pyus who had worshipped the Mother Goddess -- the Devi.
See downloaded 150816 pdf paper Bronze and Iron Age sites in Upper Myanmar: Chindwin, Samon and Pyu, by E. Moore, SOAS Bulletin of Burma Research, Vol. 1, No., 1, Spring 2003, ISSN 1479-8484 in TIL library
- Moore-Pyu-pdf<Ô> (link chk 150904).
Note: Papers in TIL SD-library can only be read on TIL computers for research.

UKT 150918: This is a note I hate to make because it involves the Hindu religions of India. Most of the scholars in Myanmarpré, including Maung (Dr.) Htin Aung, the author of ¤ Folk Elements in Buddhism
-- flk-ele-indx.htm (link chk 150918)
are under the impression that Hinduism is the religion of Hindu Trinity: Brahma (creator), Vishnu (administrator), and Shiva (the maintainer of law and order). This is the Vaishnavite Hinduism. But the dominant form Hinduism in India at present is Shaivite Hinduism, which claims that Shiva is the Omnipotent god which is comparable to YHVH, God, and Allah. In Shaivite Hinduism, Vishnu and Brahma are inferior to Shiva, and males (men) are superior to females (women). At present when a Hindu mentions "God", it is Shiva he is referring to.

We have seen an example of continuous speech, we have to know how acoustic phoneticians are studying the individual vowels produce by various human subjects -- of different age groups, of different sexes, and of different linguistic groups. Relying on a group of trained phoneticians to judge the vowels is helpful only in cursory study of a language. And you should remember that representing vowels in a diagram such as the quadrilateral of Daniel-Jones is an approximation at best.

The present-day study of vowels depends on the instrumental measurement of sound waves. A human subject is asked to "sing" a vowel and his voice-sound is recorded. From the analysis, quantities known as Formants are derived. The quantities are measured are F1, F2, and F3. Below is a comparison of vowels in F1 and F2. Again an excerpt from:
How sound is produced and heard [former hv6.htm] - snd-hear.htm (link chk 150816).

The fig below shows that using formants differentiate the vowels of {o}, {AW}/{AU}, and {au:} would not be easy. These vowels are of interest to my friend from MLC, U Tun Tint and me, because MLC transcribes the Bur-Myan {au:} /[o]/ and {o} as /[ou]/.

When I told him that in Romabama, the transliteration for is {o}, he said "that's how a man on the street {lûm:pau-ka. lu} would do it." And he is right! It is usual for male Burmese friends of the same age to address each other using the prefix {ko} (such as how I address him -- {ko htwun: tïn.}). If I were to write to him in English, I would address him as Ko Tun Tint. The explanation for how this confusion had come about is on the way the English vowels /o/ and /ɑ/ are generally pronounced. The first three formants for /o/ and /ɑ/ are quite similar, and when we pronounce {au:}/ {AU:}, foreigners might heard it as /o/. But to us, they sound as /ɑ/, and hence the Romabama transcription is {au:}.

Above, I have said to get a fuller understanding on /ɔ/ (U0254) known as "Open-O", which could be {AU}/{AW}, I need to study Mon-Myan. The only reliable way I use are the study of formants of both Bur-Myan speakers and Mon-Myan speakers. In the above figure, you can that the formants of /ɑ/ and /ɔ/ are so alike that a phonetician is bound to get confused. Since measuring formants is beyond my means at the present: all I can do is to study the Mon-Myan further.

Contents of this page

Voice quality (VQ)

Refer to Human voice, Phonetics and Phonology - indx-HV.htm (link chk 150816),
and proceed to Voice quality [former hv7.htm] - voice-qual.htm (link chk 150816)

Refer also to: http://en.wikipedia.org/wiki/Larynx 070909, 150820
See also The Phonetic Description of Voice quality, John Laver, Univ. of Edinburgh. Cambridge Univ. Press, First published 1980. ISBN 0 521 231 760. The html version is in TIL library.

Definition of VQ: Perceived characteristic ‘acoustic colouring’ of voice, derived from a variety of laryngeal and supra-laryngeal features that are not unique to one individual but form clusters of identifiable voice types. Examples of voice types:

• Whisper voice • Modal voice • Breathy voice • Pressed voice
• Creaky voice • Tense voice • Harsh voice • Nasal voice

UKT 150820: From time to time, I came across very neat sets of papers on the internet. Though I would like to give credit to the authors concerned, I am unable to do so. I can only give the source. This section is written based on two such sources. They have been downloaded, reformatted to suit TIL requirements, and are available in the TIL library.

• EGG and VQ, http://www.ims.uni-stuttgart.de/phonetik/EGG/page10.htm 080107, 150821.
The whole set of these papers have been downloaded and are in TIL library. I have gone through Chapter 1, from which the following sound files are taken.

Those who have just come across this page (file) might be wondering what the little figures in blue (given below) are. They are highly stylized drawings of the muscles of the larynx used for opening and closing the glottis.

• Whisper - www.ims.uni-stuttgart.de/phonetik/EGG/whisper.wav 101030
downloaded whisper<)) (link chk 150924)
Whisper sound quality is produced through turbulences generated by the friction of the air in and above the larynx with vocal folds not vibrating. Apart from the rather seldom linguistic uses, whisper is widely used paralinquistically to signal secrecy and confidentiality.

• Modal voice - www.ims.uni-stuttgart.de/phonetik/EGG/modal.wav 101030
downloaded modal<)) (link chk 150924)
The neutral mode of phonation is modal voiced phonation. The modal phonation of a male speaker occurs at an average of 120 Hz, while for a female speaker it is approx. 220 Hz.

• Breathy voice - www.ims.uni-stuttgart.de/phonetik/EGG/breathy.wav 101030
downloaded breathy<)) (link chk 150924)
It is normally regarded as a compound phonation type (voiceless +modal). Vocal fold vibration is inefficient and, because of the incomplete closure of the glottis, a constant glottal leakage occurs which causes the production of audible friction noise. The vibration frequency is just below the value of typical modal voice.

• Creaky voice - www.ims.uni-stuttgart.de/phonetik/EGG/creak.wav 101030
downloaded creak<)) (link chk 150924)
Creak phonation (also called vocal fry) is also produced with vibrating vocal folds but at a very low frequency.

• Harsh voice - www.ims.uni-stuttgart.de/phonetik/EGG/harsh.wav 101030]
downloaded harsh<)) (link chk 150924)
It is due to the very strong tension of the vocal folds (especially medial compression and adductive tension), which results in an excessive approximation of the vocal folds. When the whole larynx is subjected to this extremely high tension, the upper larynx becomes highly constricted with the ventricular folds pressing on the upper surfaces of the vocal folds, making their vibration ineffective. Harsh phonation is therefore irregular in both cycle duration and amplitude. The characteristic frequency is above 100 Hz. A lighter degree of tension is sometimes described as a tense voice.

• Falsetto - www.ims.uni-stuttgart.de/phonetik/EGG/falsetto.wav 101030
downloaded falsetto<)) (link chk 150924)
[falsetto.wav does not seem to represent speech]
[UKT 150821: When a grown up tries to imitate a child's voice, for example a nat-medium imitating the voice of a household ma'nes who died as a child, MaNhèLé {ma.nhè:lé:}, the voice is described in Bur-Myan as {nût-ka.tau þän}
- a nasal voice nasal<)) (link chk 150924)
It is probably what would be described as "falsetto"].

The frequency of vibrations in falsetto phonation is noticeably higher than in modal voice. The vocal folds are stretched longitudinally, thus becoming relatively thin. Consequently, the vibrating mass is smaller and the generated tone higher. The adduction of the folds is high and the medial compression is also strong. The glottis often remains slightly open, resulting in low subglottal pressure (due to constant glottal leakage) and the generation of the audible friction noise component.

Not all phonation types are mutually exclusive, on the contrary, some of them work together to modify phonation. Only modal and falsetto are incompatible because they use the structure of the larynx differently.

Voice quality is an effect of vocal tract anatomy, laryngeal anatomy and vocal habits.

This is illustrated in the following examples. The characteristic sound of the voice is brought about by the mode of vibration of the vocal cords and folds.

Differences in the degree and manner of glottal closure distinguish modal voice, breathy voice and whispery voice. The quality of the voice depends on the degree of tension in the larynx and pharynx, and on the vertical displacement of the larynx: a raised larynx produces a thin tense voice, and a lowered larynx a booming voice.

Perceptual importance: In English, apart from distinguishing voiced and voiceless sounds, voice quality does not make linguistic contrasts, but conveys information about the speaker. In some languages such as Gujerati [an Indian language that is of interest to me because of its Pali connection, and Jainism a contemporary of the Buddhism] and Mazatec differences in voice quality or pitch trajectory [The term " pitch" can mean many things. The term "pitch trajectory" is no better.] are used to convey linguistic meanings. Languages and dialects have characteristic voice qualities; personal voice quality enables a listener to recognize a particular individual. Furthermore, the quality of someone's voice also conveys emotions and attitudes.

Contents of this page

UKT notes

Tense/lax dimension

To get some more help from English, I have looked into tense/lax dimension of the vowels in English, /i/ (tense) & /ɪ/ (lax), and, /o/ (lax) & /ɔ/ (tense). They are very hard to tell apart. See Wikipedia: http://en.wikipedia.org/wiki/Vowel_diagram 150115. Looking further, I got more examples from J. Eulenberg/ A. Farhad, Spring Semester 2011, https://msu.edu/course/asc/232/Charts/Tense-Lax_Vowels.html 150815 .
This website listed:

tense: i , e , u , o , ɔ , ɑ , ɝ
lax: -- ɪ , ɛ , æ , ʊ , ʌ , ɚ , ə

The msu.edu website gives also

The Test
Only tense vowels can occur in a word-final stressed open syllable in English.
Mnemonics :
"We may view Ma Shaw's fur coat." (All tense vowels)
"Big Ben had cooked the butter." (all lax vowels)
Mnemonics IPA - my (UKT's) rendition from DJPD16 - 150815
"wiː meɪ vjuː mɑː ʃɔː fɝː koʊt" (All tense vowels)
"bɪg ben həd kʊked ðiː 'bʌt|.ɚ" (All lax vowels)

In my early days of the study of Phonetics, articles like the above on tense/lax had misguided me because unlike the other features of vowel quality, tenseness is only applicable to the few languages that have this opposition (mainly Germanic languages, e.g. English), whereas the vowels of the other languages (e.g. Spanish) cannot be described with respect to tenseness in any meaningful way. In discourse about the English language, "tense and lax" are often used interchangeably with "long and short", respectively, because the features are concomitant in the common varieties of English. This cannot be applied to all English dialects or other languages. See http://en.wikipedia.org/wiki/Vowel#Tenseness.2Fchecked_vowels_vs._free_vowels 150114

Based on "tense and lax" being interchangeably used with "long and short" in English, we may say all BEPS languages have this dimension with the further observation that "short" varieties are closer to the center of the vowel quadrilateral.

Because of the mix up of terms such as tense/lax and long/short, I am using the duration of the vowel measured in time you take to blink your eye. As an illustration consider the registers of /a/ to /i/ :

1/2 blk ----------- 1 blk ----- 2 blk --- 2 blk + emphasis
{a:.}/ {aa.} {a.} --- {a} --- {a:}
{è.}/ {èý.} {èý} {è}
{é.} ----------- {é} -------------- {é:}
{i.} ------------- {i} ---------------- {i:}

{pi.} ------------ {pi} --------------- {pi:}

The above are all free vowels. The checked vowels are those where the checking is done by killed-consonants. An example of checked vowel :