Update: 2015-12-23 10:35 PM -0500


Introduction to Romabama

- a transcription-transliteration system
for BEPS (Burmese-English-Pali-Sanskrit) languages


by U Kyaw Tun, M.S. (I.P.S.T., U.S.A.), Daw Khin Wutyi, B.Sc., and staff of Tun Institute of Learning (TIL). Not for sale. No copyright. Free for everyone. Prepared for students and staff of TIL Research Station, Yangon, MYANMAR :  http://www.tuninst.net , www.romabama.blogspot.com

index.htm | Top

Contents of this page

Rule 04. Silent e and <e> as part of digraph <ei>
Rule 05. Killed consonants
Rule 06. Special Conjuncts: {kn~si:} vowel-sign and repha
Rule 07. Fossilized killed consonants
Rule 08. Non-alphabetic characters: ~ (tilde) for {a.t} and {paaHT-hsing.}
Rule 09.
Extension of Myanmar akshara row 2 to accommodate medials and affricates


UKT notes

Contents of this page

Romabama Rule 04 

Silent e and <e> as part of digraph <ei>

e without diacritic (the "silent e" aka the "magic e") will be used occasionally for sounds of vowels followed by "killed" consonants. This is equivalent to split vowels in both

Bangla-Bengali ো (U09CB) and ৌ (U09CC), and
Bur-Myan {au:} and {au}.
The English-Latin "silent-E" or the "magic-E" may be looked upon as a split-vowel similar to those in Bangla-Bengali and Bur-Myan. Thus, into the split vowel, < i--e>, the consonant <d> is dropped to give <ide> in <tide>.

However, since both IPA and in Skt-Dev, do not use split vowels , I have avoided them in Romabama.

Silent e usually obscures the end sounds. For instance that the ending in <kate> is a non-nasal <t> sound and that <kane> ends in <n> a nasal sound, is not obvious. Whether the ending is a non-nasal or a nasal is important in Bur-Myan.

{kait} {kate} /keɪt/
{kain} {kane} /keɪn/
{lain} {lane}  /leɪn/

Note: Similar to the Vowels, Nasal endings can be realized in three registers: 

{kain.}, {kain}, {kain:}
{lain.},  {lain},   {lain:}.

however, an <e> forming part of the peak vowel is not to be confused with the silent e.

{keik} /kaɪk/ and {leik} /laɪk/
  -- the <e> present here is part of the peak vowel-digraph <ei>. It is a monophthong.

Do not forget that nasal endings can be realized in three registers: 
  (Remember {ng} stands for IPA / ŋ/ and that <g> is silent.
  Noting that {kn}, there would be no confusion if we were to drop the <g> from {ng}.)
{keing.}, {keing}, {keing:}
  {kein.}, {kein}, {kein:}
{leing.}, {leing}, {leing} 
  {lein.}, {lein}, {lein}
  -- the <e> present here is part of the peak vowel-digraph <ei>.
(Contrast with {kn.}, {kn}, {kn:}.)

The absence of a letter standing for the sound of /ŋ/ is one of un-surmountable problems of transliteration. 

{kauk} -- here <au> is the peak vowel-digraph. It is not a diphthong: it is a monophthongal digraph.


Contents of this page

Romabama Rule 05 - Killed consonants (plosive-stops & approximants)

- 150413, 150919, 151223:

UKT 150919: Though {a.} is generally held to be a vowel, since it has been included in the Myanmar consonantal akshara matrix together with approximants, we might treat it as any other consonant. Based on this argument, Romabama will allow {a.}-killed}. This is in conformity with Mon-Myan.

In the canonical structure, CV, of Bur-Myan syllables we find the killed consonants in the coda position. With the nasals, the problem becomes more complex because we have to deal with 3 pitch-registers. In the following examples, I have chosen for the onset-consonant, a typical Bur-Myan medial which is difficult for the English and Hindi speakers to pronounce.

{kyn} {kying} : nasal ending - note how g has been eliminated by changing the peak vowel from i to
{kyi} : nasal ending
{kyiN} : nasal ending
{kyn} : nasal ending
{kym:} : nasal ending

{ky} : approximant ending
{ky} : approximant ending

You will notice the peak or nuclear vowel changing in the above examples. My choice of the nuclear vowel may look artificial, but I choose it to reflect the Bur-Myan pronunciation as much as possible. However, my choice is not tenable for Mon-Myan words. I am only beginning to learn Mon-Myan and until I become confident with it, please take my choice of nuclear vowel for Bur-Myan words only.

Contents of this page

Romabama Rule 06 :

- UKT 150413, 151223:

Row 1 - column 5, {nga.}-akshara is present only in Asokan (square-shaped) and Bur-Myan (circularly-rounded). In Mon-Myan it has a hanging {ra.} as {nga.}-Mon. Words and syllables beginning with {nga.} as the onset and {kn:si:} in the coda, are very common in Bur-Myan and even in Nwari showing that it is probably common in Tib-Bur languages untainted by IE. Since Nwari is a Tibeto-Burman language, albeit under heavy influence of Indo-European Sanskrit, we should expect to see {nga.} ङ as the onset. This is what we found in the words below. Incidentally, the word for fish in Burmese is the most difficult for the Europeans and Hindi-speakers. It is almost the same in Newari and Burmese. -- UKT130116 .
See http://en.wikipedia.org/wiki/Nepal_Bhasa 140511.

गोङ्ङ 'cock' --> {gaung:nga.}
फोङ्ङा 'pillow' --> {hpaung:nga}
ल्होङ (= ल ् ह ो ङ) 'fat' --> {l~hau:nga.}
  vs. गाइसी 'thin' --> {ga-I.i}
ङा 'fish' : {nga} (long vowel) cf. Bur-Myan {nga:} (emphatic)


{kn:si:} {king:si:} /{kin: si:}/ - n. ortho. miniature symbol of devowelized nga superscripted on the following letter. -- MED2010-016

Note how  g has been eliminated by changing the peak vowel from i to

Compare the way in which the two words {hsing-kan:} and {n~kan:} are written. The first is written horizontally, but the second with the {kn~si:} {king:si:} (literally: "centipede-ridden") sign . There are two cues in Romabama to show that a {kn~si:} {king:si:} is involved: use of umlaut over the peak vowel e.g. (Alt0239) and ~. The {kn~si:} {king:si:} is actually not a conjoined sign and may be written horizontally. It is usually found in words derived from Pali and Sanskrit, e.g.:

Skt: कुङ्कुम [kuṅ-kum-a] n. saffron (a kind of crocus) - Mac069c1
Skt: कुङ्कुमत kuṅkumati f., सामान्य कुङ्कुम sāmānya kuṅkuma m. saffron Crocus sativus -- SpkSkt
Pal: {kon~ku.ma.} n. saffron, Crocus Sativus - UHS PMD0323
Bur: {kon-ku.mn} n. saffron, Crocus sativus - MED2006-024

Caution: There is an {a.t} that is not exactly a {kn~si:}, yet the consonant under it, is neither a horizontal conjunct as in Tha'gyi: {~a.}, nor a vertical conjunct such as {n~na.} & {k~ka.}. Such an {a.t} is found in:

{kywun~noap.} - MED2006-049
{yauk~kya:} - MED2006-384 

In {kywun-noap} there is only one Na'ng {na.ng} {na.}, and in {yauk-kya:} there is only one {ka.kri:} {ka.}. Yet, I am treating them as conjuncts with a ~ in between.



Sanskrit repha becomes a same-letter conjunct in Pali, e.g. धर्म dharma (= ध र ् म ) becomes  धम्म dhamma (= ध म ् म ). This change will be represented as {Dar~ma.} --> {Dm~ma.} . However, a killed {ra.}-killed in the middle of a polysyllable is not conveniently shown, and Romabama has to use a {kn~si:}-form: {Dar~ma.}

Contents of this page

Romabama Rule 07 :

Fossilized killed consonants.

(Based on personal communication with U (Dr.) Tun Tint, formerly of MLC)

There are 4 fossilized characters dating back to the 13th century:


The derivation of {rw}/{yw.} is illustrating. In the Pagan period (11th century to the 13th) and a few centuries after, the vowel {tic-hkaung:ngn-ya.t} had existed, but it has given way to {a.w-hto: wa.hsw:}. The changes have been:



Contents of this page

Romabama Rule 08 :

non-alphabetic characters

UKT: latest 150413, 151223:

{poad-hprt} (instead of 'comma') - /
{poad-ma.} (instead of 'period' or 'full-stop') - //
'period' or 'full-stop' and 'colon' are used for pitch-registers (formerly called "tones").
  They are equated to IPA suprasegmentals.
  e.g. {a.} [ă] ; {a} [a] ; {a:} [aː]

'hyphen' for separating syllables in the same word
"middle dot" (Alt0183) will be used occasionally to show that {a.} is to be pronounced as /ə/,
  e.g. {ani}.


~ (tilde) : ligature sign

The most common non-alphabetic ASCII character in Romabama is the ~ (tilde) to show a ligature of two consonantal  consonants

vertical: {paaT.hsing.}, e.g. {k~ka.} - not pronounceable
horizontal: {paaT.tw:}, e.g.  {~a.} - not pronounceable
  Note: {paaT.tw:} is my coined word for the horizontal conjunct

Essentially ~ (tilde) is used to show the hidden {a.t} over the first member of the conjunct, meaning that the character has lost its inherent vowel.

The need to show the tilde in Romabama is exemplified in transcription of Skt~Dev to Romabama, e.g. {ak} & {ak~}

Skt: अक्ष [ aksh ] - m. die for playing. -- Mac002c1
 = अ क ् ष --> {ak~Sa.}
Pal: {ak~hka.} '6-face dice' - UHS-PMD0005


Hidden ~ (tilde) : Special Skt-Dev Conjuncts

There are two special conjuncts in Skt-Dev which are commonly used, which I am calling Pseudo-Kha and Pseudo-Za . They are formed as:

Pseudo-Kha : क ् ष --> क्ष  {kSa.} /kə.sa/

Pseudo-Za : ज ् ञ --> ज्ञ  {za.} /zə.sa/

If you go by strictly Bur-Myan akshara rules, both are not pronounceable because of the viram. However, you will have pronounce them in Skt-Dev, for which you will have to include a schwa. These types of conjuncts are known in Mon-Myan as "aksharas with hangers-on" . Shown below are the first four I ran into: {na.hsw:}, {ma.hsw:}, {la.hsw:}, & {wa.hsw:}.

If we were to go by this description, we will have to call Pseudo-Kha क्ष  {kSa.} as {ka.Sa.hsw:}, and Pseudo-Za ज्ञ  {za.} as {za.a.l:hsw:}. In Mon-Myan, we meet {ka.a.kri:hsw:}.


Three-dot Tamil visarga ஃ {:.}.

The first time I notice a different use of visarga {wic~sa.pauk} (which I usually shorten to {wic~sa.}) was in Gayatri Mantra as {na:.} नः . Later I found it also in Mon-Myan. Since, I have no way to represent it, I have to borrow from Tamil visarga. It is represented in Romabama as {:.}. It signifies a very short vowel-duration of 1/2 eye-blk.


Parentheses ( )

The commonly used parentheses ( ) will be used by Romabama since it has been adopted as part of Bur-Myan.

Contents of this page

Romabama Rule 09 :

Extension of Myanmar akshara row 2 to accommodate medials

- UKT 150409, 151223

Though Bur-Myan (and Pali-Myan) akshara matrix is strictly for base consonants, Romabama has to include the medial consonants {kya.}, {hkya.} and {gya.} into row 2, to bring it in line with Pali-Latin akshara matrix.

Pal-Latin following the lead of Eng-Lat & Skt-Dev (both being IE) uses Affricates in row 2. Because Romabama uses Broad transcription (or phonemic transcription), capturing only enough aspects of pronunciation, and meaning as the key factor, I can relate Pal-Myan to Skt-Dev as:

row #2 as plosive-stop: {sa.} भ , {hsa.} छ , {za.} ज , {Za.} झ ,
row #2 as affricate:       {kya.} ,  {hkya.}     {gya.}  , {zya.} & {za.}

In the above comparison, I have left out the nasal Nya'l {a.} ञ - to make room in cell r2c4 for Mon-Myan & Skt-Dev. I have also left out the IPA and IAST transliterations to avoid unnecessary confusion.


Contents of this page

UKT notes


Contents of this page

End of TIL file