Update: 2012-11-27 06:50 AM +0630


Romabama on Typewriter


U Kyaw Tun, M.S. (I.P.S.T., U.S.A.), Deep River, Ontario, Canada. Not for sale. No copyright. Free for everyone. Prepared for students of TIL Computing and Language Center, Yangon, MYANMAR .

index.htm |Top

Contents of this page

English consonants
  Manner of articulation
  Place of articulation -- POA
  Table of English consonants
Names of the Myanmar consonants
"Killed" Consonants
Ligates or consonant clusters
UKT notes

Contents of this page

English consonants - IPA

Before we go into the English consonants, let's say something about articulation -- the act or manner of producing a speech sound especially a consonant. First we make a distinction between the manner of articulation and the place of articulation (POA).

UKT: Being a Burmese, born in Burma in the 1930s, and educated in Burma, I am "writing-biased" instead of "speaking biased". In other words, whenever I think of a language -- Burmese, English, French, and others -- I think from the way the language is written rather than how it is spoken. And therefore, whenever I think of "consonants" and "vowels", I tend to think in terms of the letters of the alphabet or aksharas. With this in mind, when I think of articulation, I have to remind myself that the "consonants" are those in the spoken language, and therefore the glyphs  representing these consonants would include "letters" not included in the English-Latin script such as /θ/ and /ʧ/ of the IPA.

The following is based on: Online Phonetics Course, Department of Linguistics, University of Lausanne, Switzerland. http://www.unil.ch/ling/english/phonetique/table-eng.html | http://www.unil.ch/ling/english/index.html

The manner of articulation is defined by a number of factors:
whether there is vibration of the vocal cords: voiced (vd) vs. voiceless (vl) -- or sonant vs. surd;
whether there is obstruction of the air stream at any point above the glottis (consonant vs. vowel);
whether the air stream passes through the nasal cavity in addition to the oral cavity (nasal vs. oral);
whether the air stream passes through the middle of the oral cavity or along the sides (non-lateral vs. lateral).

The place of articulation is the point where the air stream is obstructed. In general, the place of articulation is simply that point on the palate where the tongue is placed to block the stream of air.

The place of articulation can be any of the following:
the lips (labials and bilabials),
the teeth (dentals),
the lips and teeth (labio-dentals -- here the tongue is not directly involved),
the alveolar ridge (that part of the gums behind the upper front teeth -- alveolar articulations),
the hard palate (given its large size, one can distinguish between palato-alveolars, palatals and palato-velars),
the soft palate (or velum -- velar articulations),
the uvula (uvulars),
the pharynx (pharyngeals),
the glottis (glottals).

The tongue plays a very important role in pronunciation and the sound produced depends on where in the mouth the tip of the tongue is touching (or not touching), and how the tongue is shaped.

Table of English consonants: The regular English alphabetical table (a, b, c, d, e, f, g, ... x, y, z ) tells us next to nothing about how the consonants (and vowels) are related to articulation, and in order to overcome this drawback the International Phonetics Association (IPA) classifies the consonants into groups such as plosives (or stops), affricates, fricatives, nasals, and approximants (including laterals). Each group is subdivided when applicable into bilabial, labiodental, dental, alveolar, post-alveolar, palatal, velar and glottal.

The above table is from DJPD16 page x. The layout of the symbols follows the principle that, where there are two consonants in a cell which differ only in voicing, they are placed side by side with the voiceless one to the left. When there is only one symbol in a cell, it is placed in the centre.

Please make a note that DJPD16 has not included some symbols (such as /c/, /q/), those traditionally listed in the English-Latin alphabet.

The post-alveolar approximant /j/ is the equivalent of the common English-Latin letter <y> which has the Burmese-Myanmar {ya.} as its correspondent. The letter <y> behaves like a vowel in words like <by> and <my>, and because of this, it is also known as a semivowel.

The English-Latin <r> as an approximant is represented (after 2005) by some authors as /ɹ/ (U0279). The /r/ is classified as trill.

Though by tradition the transcription for Burmese-Myanmar {ka.} and {ga.} are English-Latin <k> and <g>, <k> sounds more like {hka.}. The same is found for <t> and <p> which sound more like {hta.} and {hpa.}. English-Latin does not have letters to represent sounds intermediate between the <k> and <g> ( in between <t> and <d>, and in between <p> and <b>). English-Latin speakers have difficulty to pronounce (or " hear") such sounds, and tend to dismiss them as the aspirated forms of <k>, <t> and <p>, and write them out with an <h> as digraphs: <kh>, <ht> and <ph>. The problem becomes worse with /ʧ/, which sounds like <ch> in <church>. However, it is more correct to represent /ʧ/ as {kya.} rather than {hkya.} or {cha.}.

Among the consonants, the plosives or stops (oral plosives, oral stops) are the most easily understood. In the following table, I have arranged the most prominent group of Burmese-Myanmar consonants which I have termed the "triads" for comparison to the IPA chart 2005 from Wikipedia. For the same group, the characters in columns c1 are voiceless, c2 described as "aspirated", and c3 voiced. Though the c2 characters sounded to be the "aspirated" c1s to the Western phoneticians, I hold that they are not so. Their pronunciations are very distinct to Burmese ears and can be described as between those of c1s and c3s. They are phonemes in their own rights. For comparison, I have given the Hindi-Devanagari akshara-characters.

Please note the different brackets used in the following tables:
  Burmese-Myanmar - {...}
  English-Latin or Pali-Latin - <...>
  IPA - /.../
  MLC - /|...|/

Bilabial plosives

  c1 c2 c3 c4
Romabama {pa.} {hpa.} {ba.} {Ba.}
MLC /|pa.|/ /|hpa.|/ /|ba.|/ /|ba.|/
Engl-Latin <p> - <b> -
Pali-Latin <p> <ph> <b> <bh>

MLC transcripts from MEDict 248, 298, 311, 319

Alveolar plosives

Under alveolar plosives, I am giving two sets of consonants. The first, r3 of Burmese-Myanmar aksharas, are regularly described as cerebrals. They are generally used for writing Pali-Myanmar. The second, or r4, are described as dentals. The reason why I am grouping both under the alveolars is because of the pronunciation. See /t d/ in Table of English consonants. Also, note the transcripts given by MLC.

  c1 c2 c3 c4
Romabama <Ta.> <Hta.> <a.> <a.>
MLC /|ta.|/ /|hta.|/ /|da.|/ /|da.|/
Engl-Latin <t> - <d> -
Pali-Latin <ṭ> <ṭh> <ḍ> <ḍh>

MLC transcripts from MEDict 160, 160, 160, 160


  c1 c2 c3 c4
Romabama <ta.> <hta.> <da.> <Da.>
MLC /|ta.|/ /|hta.|/ /|da.|/ /|da.|/
Engl-Latin <t> - <d> -
Pali-Latin <t> <th> <d> <dh>

MLC transcripts from MEDict 161, 193, 208, 216

You will notice that there are 4 characters having the same sound (at least to the Myanmar ear) in the above tables: r3c3 , r3c4 , r4c3 and r4c4 .

Velar plosives

Velars are described by older linguist, particularly of the East, as gutterals because they are articulated at the back of the mouth.

  c1 c2 c3 c4
Romabama <ka.> <hka.> <ga.> <Ga.>
MLC /|ka.|/ /|kha.|/ /|ga.|/ /|ga.|/
Engl-Latin <k> - <g> -
Pali-Latin <k> <kh> <g> <gh>

MLC transcripts from MEDict 001, 051, 083, 090

Alveolar fricatives

This group, Burmese-Myanmar r2, is described as palatals by Pali-Latin scholars. It is also where a difference between the Burmese-Myanmar and Hindi-Devanagari is found. Though there is the orthographical correspondence, there is a great difference in pronunciations. It is the source of discrepancy between the Pali-Myanmar and the Pali-Latin (International Pali). It is regrettable that many Myanmar Pali-scholars are of opinion that the Pali-Latin pronunciations to be closer to the original Pali-sounds of Gottama Buddha and Asoka, than to the Pali-Myanmar sounds. Since there are no actual sound recordings, they and we could be both right and wrong. The "manner of sound production that had been described (very meticulously) for the production of sounds" which the opposing scholars have taken to be the evidence, are those of the Brahmins in the employ of the rulers who were classed as the {hkat~ti.ya.}. Those who think that the International Pali sounds to be more authentic fail to note that the peoples just south of the Himalayas spoke similar languages long before the Indo-Europeans (Indo-Aryans) or the Brahmins (Brahmanas) came into India. We hold that the Burmese-Myanmar pronunciations are bound to be close to those of the Buddha and Asoka.

  c1 c2 c3 c4
Romabama {sa.} {hsa.} {za.} {Za.}
MLC /|sa.|/ /|hsa.|/ /|za.|/ /|za.|/
Engl-Latin <s>   <z>  
Pali-Latin <c> <ch> <j> <jh>

MLC transcripts from MEDict 100, 129, 148, 154

Note: the above table shows the orthographical correspondence between Burmese-Myanmar and Hindi-Devanagari. The r2c1 Burmese-Myanmar {sa.} (MLC /|sa.|/) corresponds to च (U091A Devanagari letter Ca). And, r6c5 {tha.} (MLC /|tha.|/) corresponds to स (U0938 Devanagari letter Sa). {tha.} behaves exactly like English-Latin <th> realized in <thin> /θɪn/ and <that> /ɪs/ (DJPD16 535.) It is noteworthy that there was a letter, thorn <>, corresponding to <th> in Old and Middle English.

It is noteworthy that English does not have c2's. These have been described to be "creaky voiced" by the some English phoneticians, and as "aspirated" by others. Note that all the c1's in the above table are pronounced at the front part of the mouth and the c3's at the back of the mouth. The c2's are pronounced somewhere in the middle.

  c1 c2 c3 c4
Bur-Myan -
Romabama {kya.} {hkya.} {gya.} -
MLC /|kja.|/ /|cha.|/ /|gja.|/ -
Engl-Latin - <ch> <j> -
Pali-Latin <c> <ch> <j> -

MLC transcripts from MEDict 026, 066, 088, --
It should be noted that r1c4 {Ga.} cannot take on any medial-formers.

English-Latin <ch> as in <church> /ʧɜːʧ/ (US) /ʧɝːʧ/; and, <j> as in <judge> /ʤʌʤ/.
Win XP character map names for Devanagari letters are: Ca, Cha, Ja, and Jha.

In the section on British English consonants in DJPD16 p. xi, it is stated that "When /l/,/j/, /w/ or /r/ immediately follow /p , t , k/, they are devoiced and are pronounced as fricatives."

In Burmese-Myanmar /p/ , /t/ , /k/ can be followed by:
/j/ (corresponding to English <y>) -- forming the medial {ya.ping.}
/r/ -- forming the medial {ra.'ris}
/l/ -- forming the medial {la.hsw:}
   -- no longer used in modern Burmese-Myanmar, except in Tavoy (according to U Tun Tint)
/w/ -- forming the medial {wa.hsw:}

Note: Conjoined characters formed from two {la.} are in use. Whether they are to be classed as medials (which have pronunciations of their own) is for me not to decide. However, since they do not occur in the onset, they are not really medials.
  {mail~la} /|mein la|/ - n. 1. somber colour; dusky colour -- MEDict359

No strict comparison can be made between the English <l> /l/ and its closest Burmese [la.] because of the absence of the 'LL' sound in English: a sound that is found in Welsh as a voiceless lateral fricative / ɬ/ (U026C). See Non-English sounds in DJPD16. That sound is present in Burmese-Myanmar as a conjunct {la. ha.hto:}. It is realised as {lha.} or as [Hla] in Burmese-Myanmar names.

Many Burmese-Myanmar syllables are of the form CV (consonant-vowel-killed-consonant). The killed-consonant () (i.e. the akshara whose inherent vowel has been killed) in the coda together with the vowel preceding it, is the 'rhyme'. In Burmese-Myanmar the vowel killer is known as {athut} (Sanskrit: virama). Of the five keyboard English vowels available <a, e, i, o, u> the most suitable to represent the inherent vowel seems to be the English 'short'  /a/ -- of course, there are bound to be exceptions. See Pronouncing the letter A in DJPD16.


Contents of this page

Names of the Myanmar Consonants

The names of Burmese-Myanmar consonants are given in the table below. The English vowel <a> has been included to give an English pronunciation.

Myanmar akshara is used for writing both the Burmese and Pali words. The traditional names are:

{wag}-aksharas (classifiables)

r1c1: {ka.} -- {ka.kri:}
r1c2: {hka.} -- {hka.hkw:}
r1c3: {ga.} -- {ga.ng}
r1c4: {Ga.} -- {Ga.kri:}
r1c5: {nga.} -- {nga.}

r2c1: {sa.} -- {sa.loan:}
r2c2: {hsa.} -- {hsa.leim}
r2c3: {za.} -- {za.kw:}
r2c4: {Za.} -- {Za.myi:hsw:}
r2c5: {a.} -- {a.kri:}; {a.} -- {a.l:}

r3c1: {Ta.} -- {Ta.than-lying:hkyeit}
r3c2: {Hta.} --{Hta.wum:B:}
r3c3: {a.} -- {a.ring-kauk}
r3c4: {a.} -- {a.r-mhoat}
r3c5: {Na.} -- {Na.kri:}

r4c1: {ta.} -- {ta.wum:pu}
r4c2: {hta.} -- {hta.hsing-htu:}
r4c3: {da.} -- {da.htw:}
r4c4: {Da.} -- {Da.auk-hkreik}
r4c5: {na.} -- {na.ng}

r5c1: {pa.} -- {pa.sauk}
r5c2: {hpa.} -- {hpa.U:htoap}
r5c3: {ba.} -- {ba.htak-hkreik}
r5c4: {Ba.} -- {Ba.koan:}
r5c5: {ma.} -- {ma.}

{a.wag}-aksharas (non-classifiables)

r6c1: {ya.} -- {ya.pak-lak}
r6c2: {ra.} -- {ra.kauk}
r6c3: {la.} -- {la.}
r6c4: {wa.} -- {wa.}
c6c5: {tha.} -- {tha.}

r7c2: {ha.} -- {ha.}
r7c3: {La.} -- {La.kri:}
r7c4: {a.} -- {a.}

You will notice that there are 4 characters having the same sound (at least to the Myanmar ear) in the above list: r3c3 , r3c4 , r4c3 and r4c4 . I have difficulty in assigning ASCII English letters to them until I decided to include the symbols (Alt0240 - similar in shape to Greek small letter Delta), and (Alt0208 - similar in shape to Latin capital letter D). and are Latin small and capital letters ETH which resemble the English letters D in appearance.

Contents of this page

"Killed" Consonants

-- consonants under {athut} (virama)

Myanmar script, like Devanagari, an abugida in which a consonant character already contains a vowel (usually the sound /a/, but in some cases as // or /ə/). The presence of the inherent vowel /a/ in the consonant characters is the main difference between the Burmese-Myanmar and English-Latin. For instance, when you are writing Romabama {ka.} for Burmese-Myanmar, "ka" stands for the English-Latin /k/+/ə/ or /k/+// or /k/+/a/. But definitely NOT for /k/+/ɑ/. /k/+/ɑ/ is the sound of Burmese-Myanmar {kau:}.   See English [a] and Bama [a] . When you are writing Romabama you will have to keep this fact always in your mind. See "Killed" consonant aksharas

Contents of this page

Ligates or consonant clusters

From the point of pronounceability, there are two kinds of ligates: those that are mute, and those that are pronounceable.

It has been noted in the section on Devanagari consonants that two consonants may be tied up or ligated. (See my note on the word " ligate".) e.g. two {ka.} consonants in a vertical ligature {kka.}. In such a ligate, the upper consonant is usually a "killed" consonant -- that is, it has lost its inherent vowel. On the other hand, the lower consonant is always a regular akshara-consonant. Thus, in the Burmese-Myanmar word for "university" {tak~ka.thol}, the ligate is {kka.} and it is mute. The upper is a killed consonant and can be represented as {k}. To explain it further, we write the word for "university" {tak~ka.thol} in the expanded form {tak-ka.thol} (Note ~ has been changed to - to show the absence of ligature). The first syllable is {tak} in CV form. The {ka.} in this syllable is a killed consonant and it has an {athut}  over it. In the short form of the word, though {ka.} has been killed, and the virama is hidden.

Pronounceable ligates are formed by the use of {ya.} as {ya.ping.}, {ra.} as {ra.ris}, {wa.} as {wa.hsw:} and {ha.} as {ha.hto:}. Though all of them are pronounceable, sometimes a ligate is just difficult to pronounce, and we ended up saying it as two consonants, e.g. {tya.}. Because of its difficult pronunciation, {tya.} is pronounced as {ta.ya.}. From this we get the word for "fine arts"  {pan-taya}. However, many Burmese-Myanmar mistakenly pronounce {pan-taya} as {pan-tra}.

Contents of this page


The following is based on DJPD16.

The English syllable (in IPA) is generally made up of three parts: an onset, a peak, and a coda and is of the form CVC. The peak and the coda constitutes the rhyme (or rime) of the syllable. The word "coda" comes from a term in music meaning "a passage at the end of a movement or composition" that brings it to a formal close [Italian from Latin "cauda" tail -- AHTD]. English allows up to four consonants to occur in the coda. In the following English words, the coda has been underlined:

<sick> /s ɪk/
<six> /s ɪks/
<sixth> /s ɪks θ/
<sixths> /sɪks θs/

The central part of a syllable is almost always a vowel, and if the syllable contains nothing after the vowel it is said to have no coda (zero coda). e.g.:

<bough> /ba ʊ/
<buy> /ba ɪ/

Note that in the word <bough>, <ough> constitutes a vowel and is represented phonemically by /a ʊ/. If you look at the word <bough> by its spelling, you might think that <gh> separately forms a coda. It is not so, and the <g> is silent.

According to DJPD16, "Some languages (e.g. Japanese) have no codas in any syllables." . Though no mention has been made about Burmese-Myanmar, it might qualify to be included because it is an abugida like Devanagari. And, the effective unit of these writing systems is the orthographic syllable, consisting of a consonant and vowel (CV) core and, optionally, one or more preceding consonants, with a canonical structure of (((C)C)C)V.

It must be remembered that even though a Burmese-Myanmar consonantal character might come after the vowel, it is actually a part of the vowel because of {athut} (virama). It is precisely because of this, I have referred to the Burmese-Myanmar structure as CV.

Contents of this page

UKT notes

   An abugida, also called an alphasyllabary, is a writing system wherein the basic symbols represent a consonant plus an unmarked vowel. When a different vowel is wanted, a diacritic or some other modification is made to the sign. The sign used to indicate the vowel is dropped is called a virama, or a " vowel killer".
   The word "abugida" comes from the first few signs of the Ethiopic Amharic script, which is an example of an abugida. Devanagari is another abugida, used in India.
   From: http://www.everything2.org/index.pl?node=abugida -- For hyper-links in this note to work go on-line.
Go back abugida-b1

Contents of this page

hear and pronounce
UKT: Native English speakers who I came across in the United States, Britain, and Australia, particularly those who have never been to Burma, can not pronounce my Burmese name, Kyaw Tun, correctly. The closest they can come to is {hkyau htun:} and eventually they "christianed" me Joe Tun. This finally forced me to officially change my name (in Canada) to Joe Kyaw Tun, the name on my passport. I hate to change my name, but the consolation prize is my interest in phonetics and languages.
Go back hear-pronounce-b

Contents of this page

The word "ligate" is a verb. (ligate v. tr. ligated ligating ligates 1. To tie or bind with a ligature. -- AHTD Though "ligate" has no noun form I will be using it as a noun -- the product of tying together.
Go back ligate-b

Contents of this page

n. Linguistics 1. The smallest phonetic unit in a language that is capable of conveying a distinction in meaning, as the m of mat and the b of bat in English. [French phonme from Greek phōnēma phōnēmat-utterance, sound produced from phōnein to produce a sound from phōnē sound, voice; See bh ā- 2 in Indo-European Roots.] -- AHTD
Go back phoneme-b

Contents of this page

n. Linguistics 1. A sound that has the quality of one of the high vowels, as (ē) or (o̅o̅), and that functions as a consonant before vowels, as the initial sounds of yell and well . Also Called glide . -- AHTD
Go back semivow-b

Contents of this page

Linguistics adj. 1. Voiced, as a speech sound. n. 1. A voiced speech sound. 2. A syllabic consonant in Indo-European. [Latin sonāns sonant-, present participle of sonāreto sound; See swen- in Indo-European Roots.] -- AHTD
Go back sonant-b

Contents of this page

surd n. 2. Linguistics A voiceless sound in speech. adj. Linguistics 1. Voiceless, as a sound. [Medieval Latin surdus ( from Latin speechless) translation of Arabic (jad _ r) 'aṣamm deaf (root), surd translation of Greek alogos speechless, surd] -- AHTD
Go back surd-b

Contents of this page

The runic letter originally representing either sound of the Modern English th, as in the and thin, used in Old English and Middle English manuscripts. [Middle English from Old English] -- AHTD
Go back thorn-b

Contents of this page

End of TIL file