Update: 2016-09-18 07:54 PM -0400


Burmese Grammar 1899

Classification of Consonants according to vocal organs


by U Kyaw Tun (UKT), Tun Institute of Learning, http://www.tuninst.net
From Burmese Grammar and Grammatical Analysis by A. W. Lonsdale, Education Department, Burma, British Burma Press, Rangoon, 1899. Copied by UKT and staff of TIL . Start: 2008 Aug.

index.htm | Top

Contents of this page

Classification of Consonants according to vocal organs
Chapter II
Rows: Guttural (velar) Palatal Retroflex Dental Labial
Columns: Nasal Deep-H Tenuis
  Vowel diagrams of Daniel Jones, and I. Catford

UKT 160416: With ref. to pix on right, the Eastern way of describing the Places of Articulation (POA) of the consonants is from the interior to the lips, whereas the Western way is from the lips to the interior: 1. Velar, 2. Palatal, 3. Retroflex, 4. Dental, and 5. Labial.

Author's footnotes
Lonsdale's footnotes, which was given at the bottom of the ink-on-paper page in the original book is now given at the end of the section.

UKT notes 
Vowels in Practical Phonetics

Contents of this page

Chapter II
Classification of Consonants according to vocal organs


Now compare Bur-Myan consonants to respective ones in other scripts. It is important for the study of Bur-Myanmar, which is slightly different from Pali-Myanmar. I have made appropriate changes in the following table from F. Mason & E. Mazard:
1. Magadhi-Asokan, 2. Pal-Myan, 3. Sinhala-Lanka, and 4. Sanskrit-Devanagari: from
A Pali grammar on the basis of Kaccayano {kic~s:} - PEG-indx.htm (link chk 160413)
  - by Rev. F. Mason, 1868 
  Downloaded versions of 251 pdf pages are available in TIL SD-Library
  - Mason-Kicsi<> / bkp<> (link chk 151022) / - PDF (link chk151022)
  - Francis Mason & Eisel Mazard (馬大影), 1st distribution in 2015
  downloaded in TIL SD-Library
  - PaliGrammar-Mason-Mazard.pdf / PalGram-Mason-Mazard<> / bkp<> (link chk 160413)

To study Mag-Asok (Magadhi-Asokan), I have to write a Lakkwak similar to the one for Bur-Myan,
a Lakkwak -- lakkwak.gif (link chk 160413)
I have several sources for this task:
#1. Francis Mason's A Pali grammar on the basis of Kaccayano {kic~s:} 1868 (E. Mazard ed. 2005, 2015)
- PalGram-Mason-Mazard<> / bkp<> (link chk 160413)
#2. Ancient Scripts
#3. Cunningham's Asoka Inscriptions , 1877
- Cunningham-Asoka-inscrip<> / bkp<> (link chk 160124)
You will notice the variations between the sources, and my version is what I consider to be a "happy medium" (Chemical Engineering Jargon which nobody likes, but use because of its utility).

There eight (or nine) vowels in Asokan Brahmi with two modes of writing: vowel-letters, and vowel-signs.

Remember, Asokan was used for writing both Pali (derived from Magadhi & Sinhala) and Sanskrit speeches. Magadhi belongs to Tib-Bur (Tibeto-Burman) language group, Sinhala to Aus-Asi (Austro-Asiatic) and Sanskrit to IE (Indo-European Aryan).

I have struck out the word Brahmi, because the Brahmin-Poannars have nothing to do with it. They failed to decipher it when called upon by the Mogul-Emperor of Delhi in the 14th century. Asokan Brahmi was deciphered by an an Englishman, James Prinsep in the 19th century. Secondly, the word Aryan has been struck out because of its association with Adolph Hitler and his racist policies of 20th century.

The guide-lines I use are mine: single circle 5 pix; double circle 7 pix. The height of cell is 22 pix. Keep in mind that a consonant and the vowel-sign modifying it are not separated by a white space.



Contents of this page

013. The Consonants are further classified on quite a different plan from that explained in par.12 - ch01-3.htm
[UKT ]

The classification is based upon the various parts of the mouth by which the sounds are produced. The organs brought into play for this purpose are the throat, the roof of the mouth, the tongue, the teeth, and the lips. [UKT 160416: Top part of the pharynx into which uvular is suspended is the throat.]

Thus there are:
Gutturals throat-letters,
   {kN~HTa.za.} (fn007-01);
   UKT 160416: POA for Engl-Lat is Velar, but can very well be Pharyngeal or throat.
Palatals roof-of-the-mouth-letters ,
  {ta-lu.za.} (fn007-02);
Retroflex aka Cerebrals 
  {moad~Da.za.} (fn007-03); [Lonsdale give this group as "Linguals" or tongue-letters
  - dismissing the term "Cerebrals". However, since in his note he asserts
  the "bringing the tip of tongue backward and upward nearly to the middle of the palate",
  I have changed it to the modern term Retroflex. -- UKT 121212]
Dentals teeth-letters ,
  {dn~ta.za.} (fn007-04);
Labials  Lip-letters
  {AUT~HTa.za.} ( fn007-005).
Each of the classes into which the letters are divided is called {HTaan} 'place'.

fn007-01 Pli,  {kN~HTa.} 'the throat'; {za.} 'to produce' [UKT: 'origin']. fn007-01b
fn007-02 Pli,  {ta-lu.} 'the palate'. fn007-02b
fn007-03 Pli,  {moad~Da} 'the head'.
  Lonsdale comments "'Cerebral is the old English designation incorrectly used for {moad~Da.za}.
  It is not a suitable term for it means 'brain letter' whereas {moad~Da.za},
  although it signifies 'produced in the head ', is used to denote a letter pronounced
  by bringing the tip of tongue backward and upward nearly to the middle of the palate.
  'Cerebral' is now superseded by the more appropriate term 'Lingual'." fn007-03b 
fn007-04 Pli, {dn~ta.} 'the teeth' fn007-04b
fn007-05 Pli, {AUT~HTa.} 'the lip'. [UKT: notice the horizontal conjunct: ] fn007-05b


Contents of this page

014. The letters are distributed into different classes mentioned above as follows: --

Gutturals [velar] {kN~HTa. HTaan} - of the {ka. wag~ga.} ka-wegga or ka group, and {ha.} of the {a.wag~ga.} awegga series. With these are included the vowels {a.} and {a}. (p007end-p008begin)

UKT 081010, ..., 160413: It is said and is commonly accepted that Bur-Myan akshara is derived from a south-Indian script through the intermediary Mon script.

However, it came as a surprise to me to find that many aksharas of the Bur-Myan bear striking resemblances to the aksharas found on the pillars of King Asoka in Magadha Kingdom of ancient India - the oldest aksharas of a complete written language in India. (Please remember that Asoka came about 250 years after the Gautama Buddha.)
See: Indic languages - indic.indx.htm (link chk 160414)

Thus, my present position is that Myanmar script is derived directly from Asokan Brahmi of Magadha Mahajanapada. That language should be termed Magadhi-Asokan, and that it was known in northern present-day Myanmarpr, particularly in the Kingdom of Tagaung and other Pyu areas since the time of King Abiraja (himself from northern India long before the time of Gautama Buddha.

If that were the case how would I explain the marked difference between Asokan & Bur-Myan in r1v1 {ka.} and r4v1 {ta.} ? Would I dare suggest that Asokan was not the parent, but that, there was a much older script from which both Asokan and Myanmar are derived. In that script, the r3, the retroflex is important, and that the full circle representing perfection is the king. Look at the first three aksharas. In Asokan, r3c2 is a full circle, and r3c1 & r3c3 are facing it as if attendance. In Myanmar, they are all on pedestals, e.g. {ta.}, {HTa.}, and {a.}. I base my conjecture on my explanation of the "magic" rune known as {sa.Da.ba.wa. n:}. See Cult of the Rune in Folk Elements in Buddhism -- flk-ele-indx.htm > ch05-magus.htm (link chk 160414)

Palatals {ta-lu. HTaan} - the {sa. wag~ga.} sa-wegga or sa group, and {ya.} of the {a.wag~ga.} awegga series. The vowels {I.} and {I} are classed with these.

UKT 160416: Though r2c5 is traditionally given as Nya'gyi {a.} (which can be under a viram without breaking up as {}), there is a hidden akshara in this place in the Bur-Myan table. The hidden akshara is Nya'l {a.}, which is clearly shown in Pali-Myan table. There is no basic {a.} in Pali. What is given as Nya'gyi {a.} is not a basic akshara (which can be under a viram without being split up): it is a horizontal conjunct which breaks up into two {a.}, as in {pi~a}.

TIL now recognizes two similar glyphs differentiated from each other by their killed-forms and different spellings in Romabama: Palatal plosive-stop {sa.}/ {c} and Dental fricative-hisser {Sa.}/ {S}.

Retroflex - changed from Lingual or Cerebral - {moad~Da. Htaan} - of the {Ta. wag~ga.} ta-wegga or ta group, and {ra.} /ɹ / and {La.} of the {a.wag~ga.} awegga series.

UKT 160416: Pronounce the Retroflex with tongue curled up, with the under-side of the tip touching the roof of the mouth.

Lonsdale assertion that {ra.} /ɹ / belongs to Retroflex is controversial. However, one should not be too much concerned, since the {a.wag} consonants are recognized as Approximants with very fluid pronunciations, and Shin Narada Thera in his An Elementary Pali Grammar course (downloaded file in TIL SD-Library
- Narada-Pali<> / bkp<> (link chk 160416)) classed them as semi-vowels. Moreover, Burmese speakers cannot differentiate between Retroflexes and the Dentals given below.

Dentals {dn~ta. HTaan} - of the {ta. wag~ga.} ta-wegga or ta group, and {la.} and {a.} /θ/ of the awegga series.

UKT 160416: In rapid speech Retroflex and Dental consonants are pronounced alike. However, if you articulate carefully, they are distinguishable: /ta./, /ʰta./, /da./, /da./, /na./. Retroflex /na./ is known as  Na'kri {Na.} 'big Na' and Dental /na./ is called 'Na'ng {na.} 'young na'. There are two forms of Na'ng: the "standing one" {na.} /na./ and the "sitting one" {na.} /na./.

Labials {AUT~HTa. Htaan} - of the {pa. wag~ga.} pa-wegga or pa group, and the vowels {U.} and {U}.

UKT 121212, 160418: I am giving below two diagrams one describes the POA, usually taught in terms of rows, in English and the other in Burmese. For the moment I will not give the translations. It is also instructive to learn the POAs in column form: c1, c2, c3, c4, c5


Contents of this page

Nasals : c5-consonants

UKT 160416: So far we have been classifying the consonants of {wag~ga.} section of Myanmar akshara-matrix in terms of rows. It is also instructive in the study of phonology in terms of columns. The most outstanding is column #5 (c5) known as nasals:  {nga.}/{gna.}, {a.}, {Na.}, / {na.}, {ma.}. Since most of us are accustomed to rows, we fail to recognize them. They are important for the coda-pronunciation of the syllables and the effect they have on the nuclear aka peak vowel of the syllable.

UKT 160416: I cannot agree with without reservation the make-up of nasals.

Firstly, {nga.}/{gna.} of r1c5, has the sound of /g/ more than /n/. The pronunciation of Burmese words such as {nga:} and {ngak} can be described as {gna:} and {gnak} similar to the situation of English <sing> and <sign> with the positions of <n> and <g> changed. I base my assertion from the way we in Kungyangoan (descendants of pure Mon-Myan speakers of Pegu-dialect who had their language changed from Mon-Myan to Bur-Myan in historical times) pronounce these words in my child-hood.

 Secondly, the cell r2c5, is traditionally occupied by Nya'kri {a.}. TIL now classifies Nya'kri {a.}/ {} as an approximant similar to {ya.}/ {}, because it can be under a virama without breaking up.

You will notice that, of all the human sounds {ma.} is the easiest to produce: in fact it is the first sound of a language used by a baby learning to speak.

015. Nasals - of column c5, {nga.}, {a.}, {Na.}, and {ma.} are Nasals or nose-letters . These are already included in the above classification, but are also called nasals because, when we pronounce them, we not only commence their sounds with the organs chiefly employed in forming them, but also allow our voice to issue through the nose instead of confining it within the mouth. The Anuthwra {a.nu.wa-ra.} aka {::tn} explained in
-- ch01-3.htm > (s010-p005) under this head.

UKT 160417:

Nasals are important in Bur-Myan, because they can appear in the coda of syllables as killed consonants: {ng}, {}, {N}, {n}, {m}. For example the intrinsic vowel of {ta.} is checked by coda {n} giving all the three registers: {tn.} (1 eye-blink), {tn} (2 blk), {tn:} (2 blk + emphasis). Note: Romabama {} rhymes with <u> of English <bun> /bʌn/.

Note the vowel-change from {a.} to {} in {tn} (2 blk). The vowel-change is not the same for all {ng}, {}, {N}, {n}, {m} giving: {tn}, {ti}, {tN}, {tn}, {tm}. My discovery of vowel-change due to the coda transforms Romabama from a transliteration (which does not reflect the pronunciation) to a transcription. Romabama spellings now reflects the pronunciation. However, the vowel-change depends on the Phonology, and since phonologies of Bur-Myan is different from Mon-Myan, you can pronounce the Bur-Myan words according to Romabama spelling. Do not pronounce Mon-Myan words from Romabama spellings.

This discovery of coda affecting the nuclear vowel is the result of my study of Skt-Dev using the Sanskrit corpus or entries in A Practical Sanskrit Dictionary by A. A. Macdonell - MC-indx.htm (link chk 160417)
comparing each entry to whatever I can find in Pali-Myanmar dictionary by U Hoke Sein.


Contents of this page

Deep-H consonants: c4-consonants

-- UKT 130907, 160418

The c4 consonants are the occupants of column #4 in the akshara 5x5 matrix of 25 plosive-stop consonants, . The akshara 5x5 matrix of 25 plosive-stop consonants form a section of matrix known as the {wag~ga.}-consonants.

They show a unique manner of articulation (MOA) in Bur-Myan and Pal-Myan. As my knowledge of Skt-Dev is meagre, and that of Mon-Myan is almost zero, I cannot say for certain what their manners of articulation are.

The 5X5 akshara consonant matrix are arranged row-wise according to place of articulation (POA), and column-wise according to MOA. Beginning from the interior, they are arranged: c5 nasals, c4 deep-H, c3 voiced (vd), c2 voiceless (vl) and c1 tenuis.

The Western phoneticians, and their IPA (International Phonetic Association) recognize only 3 columns: nasals (corresponding to c5), voiced (vd) (corresponding to c3), and voiceless (vl) (corresponding to c2 which they mix up with c1). They did not pay much attention to c4 and c2, and dismissed them just by calling them aspirates. Even then what they are saying as correspondents of c1 are actually c2. Because of this failure of IPA (the Association and the Alphabet), I have to devise my own system of Romabama to serve the needs of the Asian languages such as Skt-Dev and Bur-Myan, and use IPA to relate them to Eng-Lat.

The r1c4 phoneme of velar-consonants is given in Bur-Myan as {Ga.} with IPA /gʰ/. To study this we need to look into Bur-Myan {nga.}, which according to J. M. Haswell, who have studied the Mon-Myan language - MonMyan-indx.htm (link chk 160418),  {nga.} has the sound of English <gn> in the onsets, and <ng> in the codas.

The c4 consonants are never used to check the vowel {a.}, and so even when Macdonell has given {a.Ga.} अघ as [agh-], it should not be pronounced as as such but only as /a.gʰa/ in Pal-Myan, with a recognizable /a/ showing that the word is a negation. I emphasize the fact that my knowledge of Skt-Dev is very elementary and that I cannot pass judgment of Sanskrit sound.

In my ongoing study, I find only the entries for {a.Ga.} अघ. Similar files of r2-palatals and r3-retroflex are empty. I still have to study more of r4-dentals and r5-labials.


Contents of this page

Tenuis consonants: c1-consonants

- UKT 160417

Another column of importance in Bur-Myan are Tenuis Consonants of column c1. They are not understood by the Westerners who are used to only Alphabet-Letter system of writing. They are {ka.}, {sa.}, {Ta.}, {ta.}, {pa.}.

Tenuis consonants are important in Bur-Myan, because they can become coda of syllables as killed consonants: {k}, {c}, {T}, {t}, {p}. As an example, they can check {a.} of {ta.} giving {tak}, {tic}, {tT}, {tt}, {tp}. In every case they give only one register.


Contents of this page

The Vowels

UKT 160418: This section is my presentation.


- UKT 160417:

The vowels are continuous sounds formed inside the larynx or voice box. They are modified by the tongue inside the mouth. Until the development of fiber-optic laryngoscopy, the full involvement of the larynx during speech production was not observable. However, the role of the tongue has been studied by phoneticians such as Daniels Jones early in the 20th century. [Note only after A. W. Lonsdale had written the Burmese Grammar and Grammatical Analysis - which we are studying.] In the production of the vowels, as you move along from {a} to {au} (note: I am taking the modal pitch-register only), the tongue body is changing its position according to the red ellipse shown in the figure on the left.

The Vowel diagrams
of Daniel Jones and I. Catford

To describe the red ellipse better, it is drawn in the form of a quadrilateral by Daniels Jones. The figure on the right is based on the vowel quadrilateral of Daniel Jones in which I have shown the Burmese-Myanmar vowels. However, there are other vowel diagrams, as the one by Catford (1977) reproduced by J. Laver in the Principle of Phonetics, which is in TIL collection.

   See: Teaching Vowels in Practical Phonetics: The Auditory or Articulatory Route? by Martin J. Ball, University of Ulster, http://www.phon.ucl.ac.uk/home/johnm/ball.htm 081028. See the full article in my notes: Vowels in Practical Phonetics.

You will notice that Lonsdale has included {I.} and {I} as the vowels articulated in the Palatal position. This is in agreement with Catford. It should be pointed out that the present practice is to describe the vowels separately from the POA of consonants.


Contents of this page

016. The vowel {:} is both guttural and palatal; {AU:} is guttural and labial. [UKT: note that the POAs guttural aka velar and labial are so far apart that this "vowel" cannot be articulated. I cannot understand what Lonsdale had meant.] The consonant {wa.} is dental and labial.

017. The vowels {:} and {AU}, not being used in the Pli are not included in the above arrangement ; Burmese Grammarians, however, consider {:} to be both guttural and palatal, and {AU} guttural and labial.

018. The Burmese who have followed the Ngari system [Devanagari now used for writing Sanskrit is derived directly from Ngari] as introduced by the Pli grammarians, still keep to the classification exemplified in the foregoing paragraphs, and have devised no other. (p008end-p009begin)

A very simple method of distinguishing the Burmese consonantal sounds would be to divide the Aksharas letters into Surds, Sonants and Liquids (fn009-01), thus: -

fn009-01 Surd means 'noiseless'; Sonant 'sounding'; Liquid 'flowing'. For a full explanation of these terms, see Mason's English Grammar, pars. 13 and 15. fn009-01b

(s018end-p009end) - the next page is Chapter III.

Contents of this page

Author's footnotes

Lonsdale's footnotes, which was given at the bottom of the ink-on-paper page in the original book is now given at the end of the section.

Contents of this page

UKT notes


UKT: Wikipedia does not list suprasegmental -- 080315. However, it does say something about it in Prosody (linguistics) http://en.wikipedia.org/wiki/Suprasegmental 080318
Notice that Romabama uses IPA suprasegmentals for IPA transcriptions. Click to see what IPA has given. Or, to see the complete IPA table go to: http://en.wikipedia.org/wiki/International_Phonetic_Alphabet 080317

From Wikipedia

The prosodic domain
Prosodic features are suprasegmental in that they are not confined to any one segment; rather, they occur in a hierarchy of higher levels of an utterance. These prosodic units are the actual phonetic spurts or chunks of speech. They do not in general correspond to grammatical units such as phrases, and clauses, though they may, and both may reflect how the brain processes speech.

Prosodic units are characterized by several phonetic cues, such as a coherent pitch contour, and the gradual decline in pitch and lengthening of vowels over the duration of the unit, until the pitch and speed are reset to begin the next unit. Breathing, both inhalation and exhalation, only seems to occur at these boundaries where the prosody resets.

UKT: Only some suprasegmentals are of interest to Romabama, e.g. {a.} /ă/; {a} /a/; {a:} /aː/.

Go back supraseg-note-b

Contents of this page

Vowels in Practical Phonetics

by Martin J. Ball, Teaching Vowels in Practical Phonetics: The Auditory or Articulatory Route? . University of Ulster. http://www.phon.ucl.ac.uk/home/johnm/ball.htm 081028
UKT: I haven't taken the figures given by M. J. Ball since they are essentially the same as those already on this file.

Vowel description has traditionally differed from consonant description. This means that teaching the production of vowels and consonants in practical phonetics requires different techniques. Unfortunately, this means that students have to master both approaches to their studies.

Consonants are described and classified according to their production: manner of articulation (e.g. stop, fricative); place of articulation (e.g. alveolar, palatal, velar, uvular); and voicing. Students, therefore, can use these labels to aid in learning the articulation of consonants.

The Vowel Area and Cardinal Vowels :
Vowels, on the other hand, have traditionally been taught via the Cardinal Vowel System. This system is based on a set of auditory reference points, and vowels are described in terms of how close they are to these reference points. Cardinal Vowels, therefore, have to be learnt by students auditorily: through imitating models, not by learning their articulation.

In the most recent revision of the International Phonetic Alphabet (1993), the vowel diagram was re-arranged (see Figure 1)[UKT: essentially the same as the one I have redrawn by UCLA given on the right.]. All unrounded vowels are displayed on the left of each point. All rounded vowels are displayed on the right. A full set of central vowels is provided, and symbols for lax vowels are added to the diagram.

Vowel Area and Cardinal Vowels:
The vowel quadrilateral is, in fact, quite divergent from the actual shape of the vowel area (see Figure 2) [UKT: essentially the same as the one on the left.]. Catford (1977), among others, has suggested a diagram closer to physical reality could be adopted. This could allow articulatory descriptions of vowels, similar to those used for consonants. In order to produce a diagram closer to the vowel area, the angled corners of the Cardinal Vowel diagram need to be abandoned, and a chart nearer to the ellipse shape in Figure 2 created.

An Articulatory System
Vowels are the next most open articulation type after approximants and fricatives. This means that close vowels can easily be linked to the palatal, velar, uvular and pharyngeal places of articulation. Students learn their production by moving the tongue slightly between consonantal and vocalic versions at each place.

An articulatory system follows the vowel area more closely, and this means that [i, i, u, o, a, a] are all located on the upper periphery. Other vowels are labelled as being close-mid, open-mid or open in relation to one of the places of articulation. Due to the shape of the diagram, the lower left corner vowel [a] is both an open palatal and an open pharyngeal vowel.

Advantages and Disadvantages to an Articulatory System
Only one set of articulatory labels need to be learnt, and the same method of learning sound production can be applied to consonants and vowels. Further, the vowel diagram is closer to the vowel area.

However, only production of the upper periphery vowels is easy to learn, as it is unclear how one learns the values of close-mid, open-mid and open. An articulatory system aids learning vowel production, but not description: description of both consonants and vowels is an auditory task - as is using Cardinal Vowels.

Phonologically, three-vowel systems ([i, a, ul) plot better on the CV system, as they appear clearly as 'extreme' vowels. Languages having high, mid, and low vowel phonologies group their vowels as in the CV system rather Um the polar co-ordinate system. Also it is difficult to arrange central vowels on the polar co-ordinate diagram, and show their relation to peripheral vowels.

Catford (1977) points out reasons why the CV system appears more natural: if we plot vowels acoustically (FI by F2), the resultant diagram closely resembles the CV chart. He also notes that the muscle systems used to move the tongue within the vowel area give us proprioceptive feedback that high versus low, and front versus back are natural classes: as shown in CV diagrams.

An articulatory system certainly could make the learning of vowel production easier. The system also brings consistency between vowels and consonants. However, it does not help in vowel description; and has phonological problems. It is doubtful, therefore, whether a switch in teaching vowels in practical phonetics is warranted.

References (cited by M.J. Ball)
Catford, I. (1977) Fundamental Problems in Phonetics. Edinburgh: Edinburgh University Press.
IPA (1993) Council actions on the revisions of the IPA. Journal of the International Phonetic Association, 23, 32-34.

Go back vow-pract-note-b

Contents of this page

End of TIL file