Vowels

Consonants

Unlike Hindi, Gujarati no longer has a vowel length contrast for /i/ and /u/, though they are still distinguished in writing by most writers.

The vowel ऋ <r̥> in Sanskrit originally represented a syllabic liquid like /ɹ̩/, in modern Bengali this is pronounced as a sequence /ɾu/ in words directly derived from Sanskrit, which is not a different vowel from /u/, but the original Sanskrit vowel letter is still indicated in the spelling.

Consonants

The first 25 letters in a typical Indic-template script consist of five series of five letters each. The five letters in a series represent voiceless unaspirated, voiceless aspirated, plain voiced, breathy voiced, and nasal consonants, in that order. The five series are velar (k row), palatal (c row), retroflex (ṭ row), dental (t row), and labial (p row).

The remaining consonant letters in the main Indic-template block consist of two series, one series of sonorants (y, r, l, v) and one series of fricatives (ś, ṣ, s, h). /ŋ/ and /ɲ/ are not contrastive in Gujarati; they only occur before the consonants in their respective series.

The dental series consonants are apical dental consonants where the tip of the tongue must touch the back of the front teeth, so alveolar coronal consonants like /t/ and /d/ in English loanwords are perceived as the retroflex series consonants instead and spelled as such. The dental component is not necessary in the case of /n/ because the retroflex /ɳ/ is generally realized as [ɽ̃] except when coming before another retroflex consonant.

The /pʰ/ sound for many speakers is realized as [f].

/ɾ/ can also be realized as a short trill, and when geminate, it is typically realized as a trill [rː].

There is typically no distinction between the palatal and retroflex sibilants /ʃ/ and /ʂ/, with both being pronounced as /ʃ/. Some speakers also do not distinguish this /ʃ/ from /s/.

Other Vowels and Vowel-like Marks

ं (anusvar) before a consonant indicates a homorganic nasal consonant from that same series. It can also be used to indicate nasalization of a vowel, especially in word-final position.

ः (visarga) indicates a soft h pronunciation after the vowel, optionally followed by a short echo of the vowel sound again. For the most part this is used only in words derived directly from Sanskrit, with a couple exceptions. It is romanized as <ḥ>.

There are two vowel letters that can be used for /æ/ and /ɑ/ ~ /ɔ/ in English loanwords:

There are also some other Sanskrit syllabic liquid vowel letters that are extremely rare, even in Sanskrit:

These letters will practically never be encountered in written Gujarati unless one is reading specifically about Sanskrit.

Main Vowels

The inherent short vowel in Hindi is often described as a shwa /ə/, though the phonetic realization is typically lower as [ɐ]. The distinction between it and its long counterpart /a:/ is typically neutralized in word-final position to [a].

The Sanskrit distinction between short-long counterparts for /i/ and /u/ is lost in Assamese, but still spelled out in the writing system where etymologically appropriate.

The vowel ऋ <r̥> in Sanskrit originally represented a syllabic liquid like /ɹ̩/, in modern Assamese this is pronounced as a sequence /ɹi/ in words directly derived from Sanskrit, which is not a different vowel from /i/, but the original Sanskrit vowel letter is still indicated in the spelling.

Assamese does not distinguish the retroflex and dental coronals as is typical of most Indo-Aryan languages, having merged both sets together into an alveolar series.

Assamese shifted the three sibilants to velar /x/, and subsequently shifted the palatal series plosives to /s/ and /z/.

Other Vowels and Vowel-like Marks

ং indicates a velar nasal /ŋ/ (in Lexx Rom romanized as ṅ same as ঙ).

ঁ indicates nasalization of the vowel.

ঃ represents /h/ after a vowel, with an optional echo of the vowel afterwards as well. is used only in words derived directly from Sanskrit, romanized as ḥ.

There are also some other Sanskrit syllabic liquid vowel letters that are extremely rare, even in Sanskrit:

These letters will practically never be encountered in written Assamese unless one is reading specifically about Sanskrit.

Bengali is written in the Eastern Nagari script, also commonly known as the Bengali script. The Eastern Nagari script is an Indic-based script that is similar to Devanagari (used for Hindi, Marathi, and other languages). Eastern Nagari script is also used to write Assamese. As with other abugida-style scripts, there is an inherent vowel with each consonant letter.

Main Vowels

The inherent vowel in Hindi is usually /ɔ/, but in some cases, it is /o/.

Bengali does not distinguish between short and long forms for /i/ and /u/, but the original letters from Sanskrit are still spelled out in the writing system.

The vowel ऋ <r̥> in Sanskrit originally represented a syllabic liquid like /ɹ̩/, in modern Bengali this is pronounced as a sequence /ɾi/ in words directly derived from Sanskrit, which is not a different vowel from /i/, but the original Sanskrit vowel letter is still indicated in the spelling.

Main Consonants

Bengali does not have a separate letter for Indic v/w, but the letter y can be modified with a in front to become w.

Bengali also modifies the retroflex voiced and breathy voiced consonants with a nuqta dot to create flapped versions:

Many speakers merge ṛh into ṛ, and some speakers additionally merge ṛ with normal r.

Other Vowels and Vowel-like Marks

ং indicates a velar nasal /ŋ/ (in Lexx Rom romanized as ṅ same as ঙ).

ঁ indicates nasalization of the vowel.

ঃ represents /h/ after a vowel, with an optional echo of the vowel afterwards as well. is used only in words derived directly from Sanskrit, romanized as ḥ.

There are also some other Sanskrit syllabic liquid vowel letters that are extremely rare, even in Sanskrit:

These letters will practically never be encountered in written Bengali unless one is reading specifically about Sanskrit.

Letters

The Devanagari script used for Hindi is an abugida-style alphabet, with an implicit short a vowel inherently part of the base consonant letter. The Lexx Rom romanization system for Hindi is largely in line with ISO Indic transliteration.

Main Vowels

The inherent short vowel in Hindi is often described as a shwa /ə/, though the phonetic realization is typically lower as [ɐ]. The distinction between it and its long counterpart /a:/ is typically neutralized in word-final position to [a].

Similarly, the distinction between short-long counterparts /ɪ/ /i:/ and /ʊ/ /u:/ is neutralized in word-final position to [i] and [u] respectively.

The vowel ऋ <r̥> in Sanskrit originally represented a syllabic liquid like /ɹ̩/, in modern Hindi this is pronounced as a sequence /ɾɪ/ in words directly derived from Sanskrit, which is not a different vowel from /ɪ/, but the original Sanskrit vowel letter is still indicated in the spelling.

/e:/ and /o:/ do not have short separate short vowel versions, but in terms of derivational morphology, in words where they reduced they often correspond to short /ɪ/ and /ʊ/ respectively.

/ɛ:/ and /ɔː/ originally represented diphthongs /ai/ and /au/, but in standard Hindi/Urdu these are now monophthongized. They are often still diphthongs in other Indian languages, so the romanization still reflects the original diphthong origin.

Short [ɛ] and [ɔ] can surface as allophones of /ə/ when two /ə/ surround the letter /ɦ/.

In Hindi, the inherent short a vowel can be deleted in certain positions, such syllable-final position. Where not pronounced, it will be left out of the Lexx Rom romanization.

Main Consonants

The first 25 letters in a typical Indic-template script consist of five series of five letters each. The five letters in a series represent voiceless unaspirated, voiceless aspirated, plain voiced, breathy voiced, and nasal consonants, in that order. The five series are velar (k row), palatal (c row), retroflex (ṭ row), dental (t row), and labial (p row).

The remaining consonant letters in the main Indic-template block consist of two series, one series of sonorants (y, r, l, v) and one series of fricatives (ś, ṣ, s, h).

/ŋ/ and /ɲ/ are not contrastive in Hind; they only occur before the consonants in their respective series.

The dental series consonants are apical dental consonants where the tip of the tongue must touch the back of the front teeth, so alveolar coronal consonants like /t/ and /d/ in English loanwords are perceived as the retroflex series consonants instead and spelled as such. The dental component is not necessary in the case of /n/ because the retroflex /ɳ/ is generally realized as [ɽ̃] except when coming before another retroflex consonant.

The /pʰ/ sound for some speakers is pronounced as [f].

/ɾ/ can also be realized as a short trill, and when geminate, it is typically realized as a trill [rː].

There is typically no distinction between the palatal and retroflex sibilants /ʃ/ and /ʂ/, with both being pronounced as /ʃ/. Some speakers also do not distinguish this /ʃ/ from /s/.

Other Consonants

Some supplementary consonants in Hindi that aren't part of the core Indic-template consonant block can be indicated through the use of a nuqta, or dot diacritic, placed underneath a base consonant letter.

Hindi modifies the letters for <ḍ> and <ḍh> with the nuqta to create their flapped counterparts. These sounds originally developed when <ḍ> and <ḍh> became flapped in non-intial position, so they do not typically appear at the beginning of a word. Due to the subsequent influx of loanwords into Hindi-Urdu after this sound shift had occurred where retroflex <ḍ> can occur in this position, there are now minimal pairs between these and the original plosive consonants.

These nuqta letters are used for sounds that entered Hindi-Urdu through loanwords, especially loanwords from Arabic, Persian, and English. They are often not written out with the nuqta, so you can often see these words just written using the base letter instead. In terms of pronunciation, many speakers do not distinguish the pronunciation of these letters from the base letter, but English-educated speakers will typically use /z/ and /f/ where appropriate in an English loanword. As mentioned above, some speakers also use /f/ across the board all the time instead of the original /pʰ/ sound. Some English loanwords /ʒ/ with can occasionally be seen with झ़, but it is not often distinguished from ज़ or ज.

The letters क़ ख़ ग़ are found in words of Arabic or Persian origin, and so they are generally more characteristic of Urdu-register pronunciation.

Other Vowels and Vowel-like Marks

ं (anusvar) before a consonant indicates a homorganic nasal consonant from that same series.

ँ (candrabindu) above indicates nasalization of a vowel. In cases where it would not fit size-wise due to another vowel mark competing for the same space, the nasalization is indicated with the ं anusvar.

ः (visarga) indicates a soft h pronunciation after the vowel, optionally followed by a short echo of the vowel sound again. For the most part this is used only in words derived directly from Sanskrit, with a couple exceptions. It is romanized as <ḥ>.

There are two vowel letters with a candra (half-moon) without the bindu dot, that can be used for /æː/ and /ɑː/ in English loanwords:

Of these, ऍ is now obsolete in Hindi, being replaced with ऐ in English loanwords where the vowel would be used (meaning that in some words ऐ can indicate æː rather than ɛ:). ॅ is still used in Marathi, though the base vowel is considered to be ॲ.

Much like the English /ɔː/ is not always distinguished from /ɒː/ or /ɑː/ depending on dialect, vowel /ɑː/ is not always distinguished from /a:/ (which itself can be realized quite towards [ɑː] in Hindi). As such, it is often without the candra as the simple आ.

There are also some other Sanskrit syllabic liquid vowel letters that are extremely rare, even in Sanskrit:

These letters will practically never be encountered in written Hindi unless one is reading specifically about Sanskrit.

Burmese is a Sino-Tibetan language related to various other Burmic languages such as the Zaiwa, Achang, Hani, Lahu, Lisu, and Yi languages of northern Myanmar and southwestern China, and more distantly related to other language branches of Sino-Tibetan such as Sinitic, Karenic, and Tibetic. Within Burmic, its closer relatives are those of the Burmish clade such as Achang and Zaiwa, and within this clade the Burmese cluster includes Intha, Danu, Taungyo, Tavoyan, Rakhine of Rakhine State, and Marma of eastern Bangladesh.

The Burmese script is the most prominent orthography using the greater Myanmar script (Mon-Burmese script), which is also used for many other languages within the borders of Myanmar, such as Mon, Shan, Sgaw Karen, Pa-O, and others. The Myanmar script is thought to have been adapted for Old Burmese from Old Mon, and it is ultimately of Indic Brahmi origin. As such, it is helpful to learn some of the basic principles underlying the Indic script model to understand how the script was adapted to the features of Burmese.

Initials
In the table below, "Ancestral Indic" refers to the value of the consonant in the broader Indic script system, based on the Sanskrit/Pali sound, which sometimes differs from the sound in modern Burmese. All consonant initials carry an unwritten inherent vowel ā (/a̰˥ˀ/).

The typical Indic script template starts out with 5 basic series of 5 letters each; these five series each form a pattern of voiceless unaspirated consonant, voiceless aspirated consonant, voiced consonant, breathy voiced consonant, and nasal initial. The five series are organized by the part of the oral cavity that engages the sound, so they are the k kh g gh ṅ (velar), c ch j jh ñ (palatal), ṭ ṭh ḍ ḍh ṇ (retroflex), t th d dh n (dental), and p ph b bh m (labial), from back of the mouth to front of the mouth. The Indic template then lists out a row of sonorants (in Burmese the y r l w) and a row of fricatives (in Burmese reduced to just s and h), then an extra sonorant ḷ used for Pali but not in Sanskrit. The Indic template letter for the short a (အ) follows after as it is used as the base initial for the glottal stop initial/null initial (commonly done in many Southeast Asian Indic orthographies).

Consonants can take a "killing stroke" ် on top of them that deletes the inherent vowel. At the end of a syllable, this indicates a coda (details in the rimes section). Word-medially, another way of deleting the inherent vowel is by creating a combined consonant letter, generally made by stacking the consonant on top of the subsequent consonant, so က + က = က္က. This is used for Indic-derived words.

ဃ ဈ ဎ ဓ ဘ
As can be seen from the table, Burmese does not pronounce these Indic breathy voiced initials (gh, jh, ḍh, dh, bh), merging them with their plain voiced counterparts. ဘ is actually the more common letter used for /b/ over ​ဗ, and ဓ is also used to some extent, but the others are uncommon.

စ ဆ ဇ ဈ
The Indic palatal series စ c,​ ဆ ch, ဇ j, ဈ jh has spirantized into modern Burmese s, hs, z, z. Some speakers do not distinguish aspirated hs /sʰ/ from /s/.

ည ဉ
The palatal nasal ñ has two forms; for initial ñ, ည is the default letter. The most common use for ဉ is as a final nasal coda ဉ်, though ည် is also used as a rime as well (details in the rimes section). ဉ is also used when this ဉ် is the first consonant in an Indic-derived stack such as ဉ္စ ñc ဉ္ဆ​ ñch ဉ္ဇ ñj ဉ္ဈ ñjh.

While ည is the default letter for ñ in modern Burmese, in some Indic-derived words, ဉ is still used as consonant initial ñ. When an Indic word has this consonant initial ñ attached after another ñ, it becomes written identical to ည, but learners need to remember which words use ည as ဉ် + ဉ as they encounter them.

ဋ ဌ ဍ ဎ ဏ ဠ
The Indic retroflex consonants are not distinguished from the dental consonants in Burmese pronunciation, but are used for etymological spelling/transcription of the sounds in Indic words where needed, so they are relatively uncommon in Burmese. The letters ဋ ဌ ဍ ဠ (ṭ ṭh ḍ ḷ) have their default stacked forms turned sideways to fit more easily: ္ဋ ္ဌ ္ဍ ္ဠ. There are also special stacked combinations ဋ္ဋ ṭṭ ဋ္ဌ ṭṭh ဍ္ဍ ḍḍ ဍ္ဎ ḍḍh ဏ္ဋ ṇṭ. Note that the glyphs for ဌ (ṭh) and ဋ္ဌ (ṭṭh) appear quite similar.

ယ ရ
ရ r merged into ယ y, but in modern loanwords it can be used to indicate the r sound. There are also some occasional Indic-derived words that preserve the r sound. This r-y merger did not occur in the Rakhine language (Arakanese).

Learners should note that the <y> sound in Burmese can be realized as [ʝ] or [ʒ] for in the speech of some speakers.

သ
Ancestral Indic သ s became Burmese <thh> /θ ~ t̪θ ~ t̪/ (part of the chain shift that allowed c to spirantize into s), somewhat analogous to how some speakers of English may lisp and pronounce the s sound as a th. This sound is often described as a dental fricative, but the fricative component of it is quite reduced in most speakers, and learners should prime attune ears to listen for something closer to a dental /t̪/ sometimes with slight affrication /t̪θ/. When two သ letters are combined together, they can become a special letter ဿ.

Consonant Voicing

The voiceless first two consonants in each of the five basic consonant series (voiceless unaspirated, voiceless aspirated) can become voiced in non-initial position (changing to the corresponding voiced consonant in that row). This is very common in Burmese, but not a universal rule, so it is something that learners should pay attention to in listening and in the romanization when they are learning new words, because Burmese orthography does not use any mark to overtly indicate this voicing change in pronunciation, unlike, for example, in Japanese orthography, where this process (known as 連濁 rendaku in Japanese) is overtly indicated with the dakuten mark ゛on the letter. So for example in the Burmese word စကား, which has two syllables စ and ကား, the ​က in the second syllable actually gets voiced from k to g. Theoretically, if Burmese orthography were like Japanese orthography, စကားﾞ could be a way to represent in the script that this consonant change happens, but since it is not explicitly written out, learners will have to simply remember when it happens or doesn't happen. There are some principles to help predict when it doesn't happen, which will be brought up in later sections.

(သှ) <dhh>
In addition to the first two consonant letters in each of the five basic series, the letter သ <thh> (/θ ~ t̪θ ~ t̪/) can also undergo voicing, changing to its voiced equivalent /ð ~ d̪ð ~ d̪/, romanized as <dhh> in Lexx Rom. As mentioned above, consonant voicing is not explicitly marked out in Burmese, so you will always just see this sound written with the same သ letter in Burmese, but some reference books for learners use သှ with a mark  ှ to indicate this changed sound (more on this mark  ှ later).

ျ ြ ွ
There are three glide consonants that are commonly paired with Burmese consonant initials:

Notice that in Standard Burmese, the ရ -r - ယ -y- merger also applies when they are glides. As in the case with initial r-, Rakhine still preserves the -r- pronunciation of this glide.

Historically, there was a -l- medial glide ္လ as well, which in Standard Burmese not only merged with the -r-/-y-, but also got eliminated in spelling, whereas the -r- glide is still distinguished in spelling. However, in some parts of the Burmese linguistic clade such as Intha and Tavoyan, this -l- medial is still spelled and pronounced.

In some Indic-derived words, you may encounter a y initial with a -y- glide ယျ; in Burmese, this is pronounced identically to single ယ.

When the -w- glide and either of the other two glides are present together on the same carrier consonant, the -w- glide is dominant and the other glide is generally not pronounced unless one decides to do so for very careful, literary, recitation-style speech.

◌ွ can also be used underneath ဗ b as ဗွ to indicate the sound /v/ (Lexx Rom v) in modern loanwords, though Burmese is inconsistent about this, so plain ဗ by itself is sometimes used for this v sound as well.

ကျ ချ ဂျ ဃျ ငျ ကြ ခြ ဂြ ဃြ ငြ
When the velar series က k, ခ kh, ဂ g, ဃ g  င ng take on either of the two -y- medial glides, they become palatalized into palatal affricates:

Of these, you are not going to run into the combinations ဃျ ငျ ဃြ ငြ are rare (ဃျ ငျ ဃြ are theoretically only needed if dealing with direct transliteration of Indic texts), but they are listed here for completeness.

Remember that the original ancestral palatal series had spirantized into the modern s, hs, z series, so that made space in the phonological system for this new series of palatalized consonants to emerge.

◌ှ
Burmese also has another letter ှ that looks like a glide but is more of a devoicing marker, equivalent to adding an initial ဟ h- in front of the modified letter:

Note that in နှ the long tail of the letter န gets folded up to make space for this ှ marker. Also note that the <hng> combination should not be mixed up with the similar-appearing ဌ (ṭh) and ဋ္ဌ (ṭṭh).

This devoicing mark causes different behavior on the two y letters ရ (ancestral r) and ယ (ancestral y), causing them to turn into a sibilant /ʃ/ <sh>. This sound also applies to a couple other idiosyncratic consonant clusters as well (to be discussed later).

This <sh> sound is by far most commonly spelled with ရှ, but some dictionaries use ယှ as the default form.

Very rarely, a writer may use ှ under the letter ဖ ph as ဖှ to transcribe the sound /f/ (Lexx Rom f) from modern loanwords. Most frequently though, Burmese will generally just use the plain letter ဖ by itself even if pronouncing it as f.

ၐ ၑ
The Burmese script also has two special consonant letters for <sh> that are only used when writing Sanskrit (they do not occur in Pali):

These two characters are practically never used in written Burmese (Pali merged them into /s/, and it is more common for Burmese to borrow Pali forms). Learners should remember ရှ as the default way to indicate <sh>.

Other letters dedicated to transcribing consonant sounds from other languages are even less common, but ဍ့ (ḍ with ring underneath) and ​ဎ့ (ḍh with ring underneath) have been documented for Hindi retroflex flaps ड़ /ɽ/ ṛ and ढ़ /ɽʱ/ ṛh.

Tones

Before getting to rimes, the tones of Burmese will be discussed, as Burmese spells some of its rimes slightly differently depending on the tone.

Burmese has four tones, a mid-low tone, a high creaky tone, a high tone, and a checked tone with a glottal stop.

Tones

The low tone is the default unmarked tone in Burmese; it is the most commonly encountered tone. It is a mid-low flat tone that can have a little bit of optional updrift towards the end.

The creaky tone is a relatively short high-pitched tone that carries some creaky glottalization, and it can have downdrift towards the end. It is the least common of the four tones in Burmese, and is generally indicated with a ring towards the bottom right of the syllable( ့ ).

The high tone is a high falling tone that can sometimes carry some breathiness. It is generally indicated with two rings to the right, the Indic visargah (း).

The checked tone is analogous to the Sinitic Rusheng tone as well as the Tai-Kadai "dead" tone (East Asian Tone Category D); this tone is only paired with syllables that end in checked coda (a glottal stop coda in modern Burmese). If a syllable does not end in a checked coda, it must only be in one of the other three tones (mid-low, creaky, or high). The checked tone is a high tone that is stopped abruptly by a glottal stop (the "uh" sound in the English uh-oh); from its starting point, it can fall in pitch a bit before it reaches its glottal stop ending.

Overall, the Burmese tone system is relatively fluid and forgiving compared to the tone systems in many other languages of East and Southeast Asia, having been described as being in a state of instability or advanced decay. In isolation, the creaky tone may sound like the checked tone to learners, but one clue to distinguish them is that they take different sets of rimes (since the checked tone can only take checked rimes). In connected speech, the creaky tone is starting to merge to some degree with the high tone for some speakers, but the two tones are still distinguished in writing.

Rimes

Burmese has seven basic vowels. The table below shows their representation in the three non-checked tones of Burmese:

Basic Vowels

For the Burmese vowels a, i, and u, the glyph for the short vowel in the Indic script template was utilized for the vowel with creaky tone, and the glyph for the long vowel was utilized for the vowel with low tone (there is no short-long vowel length distinction in Burmese).

The vowel a in the low tone is usually written as ာ, but for certain consonants, if it would look too similar to a different consonant,  a tall version ါ is used instead. This applies to the following consonants:

​ဂါ (to distinguish from က)​
ငါ (to distinguish from က)​
ဒါ (to distinguish from အ)
ဓါ (to distinguish from မာ)
ပါ (to distinguish from ဟ)
ဝါ (to distinguish from တ)

The same principle applies to the vowel ɔ, as its right-side component consists of the same shape as the vowel a.

For the vowel ɛ, a y letter with the vowel deletion marker ◌ယ် is appended to the base consonant in the low tone, while the special vowel letter is used for the high tone (which doesn't need the း), and the creaky tone is created regularly.

For the vowel ɔ, the vowel deletion marker ် is used to indicate the low tone, so a syllable without this marker has the high tone even without the  း. The creaky is created regularly.

These vowels with the null initial/glottal stop initial are generally appended to the null consonant , but in some words (mostly Indic-derived words), some of the null-initial vowels have dedicated letters.

ဧ is generally <e>, but in some words it represents <ê>.

ဦ can also take the high tone marker as ဦး <û>.

Initial Minor Syllable

Over time, many Burmese words have shifted towards sesquisyllabic structure, which is typologically common in Southeast Asia. A typical Southeast Asian sesquisyllabic (literally meaning 'one and a half syllable') word is made of a minor syllable plus a full (normal) syllable.

In Burmese, the minor syllable can only have one rime: an initial consonant plus the unstressed schwa /ə/, which in Burmese only occurs in this minor syllable. In terms of pitch height, this neutralized syllable is a lower pitch than subsequent syllables.

There is no specific letter or way to write this schwa, because it can be a result of any first syllable in a word getting neutralized and destressed into a minor syllable, and this neutralization of the syllable is not reflected in the orthography. In many cases, a first syllable with the creaky tone ā (meaning no additional vowel marker) will become reduced into a minor syllable, but any rime, even one with a final coda, can become reduced into a minor syllable. In Lexx Rom, this is indicated by <ă>. Some books for learners will use an apostrophe if wanting to mark out that the entire preceding syllable becomes a destressed minor syllable.

Destressing a syllable into a minor syllable will often also cause a minor syllable's initial to change into its voiced counterpart, if available (review the section above on Consonant Voicing). As nothing in the spelling of a word is changed with the destressing of a syllable, learners will need to pay attention to pronunciation and Lexx Rom when watching out for this change. For example, the Burmese word ပုစွန် 'shrimp' has two syllables ပု <pū> + စွန် <sun>, but when combined, not only does the စ in the second syllable get voiced from s to z due to consonantal voicing, but the first syllable gets destressed, and correspondingly the ပ changes from p to b. The final result is that ပုစွန် becomes <băzun>, quite different from the <pūsun> that the spelling would otherwise suggest.

Nasal Rimes

In Burmese, nasal codas have all merged together and boiled down to a weak uvular nasal /ɴ/ that often just surfaces as nasalization on the preceding vowel with a slight glide [ɰ̃]. The nasal coda just transcribed as <-n> in Lexx Rom, but the glyph letter representing it in the Burmese script can be င် -ng, ဉ် -ñ, န် -n, မ် -m, or ံ -ṃ, depending on the rime (ṃ is the Indic anusvara that indicates either nasalization or a nasal consonant).

The " in the table indicates the the rime is pronounced identically to the rime above in modern Burmese. Spelling should be memorized by lexical item.

Sometimes, mostly in Indic-derived words, the nasal rime င် in front of a velar consonant can take on a new miniature form ​င်္ that rides on top of the velar consonant. Its pronunciation remains the same.

Learners should note that the /a/ in the /aɴ/ rime can get fronted to result in [æɴ] for many speakers.

Irregular Coda ည်

ည် may appear to be a nasal rime coda -ññ, but it is actually a highly irregular coda letter that can indicate one of three basic vowel rimes <-i>, <e>, or <ę> depending on the word. Which vowel is used must be learned word by word, but in general, -i is considered the literary pronunciation for many words, and <ę> its colloquial counterpart.

Confusingly, in pre-modern texts ည် was also often used for the <-in> rime, but in modern orthography those words are now assigned to the half-sized ဉ်, leaving ည် still with the significant load of three basic vowel rimes.

Checked Rimes

In Burmese, all checked codas have merged together into a glottal stop coda, which Lexx Rom transcribes as -t (common practice in Myanmar, based on the English glottal stop allophone of t). The glyph letter representing this checked coda in the Burmese script can be က်, စ်, တ်, ပ် p, depending on the rime.

As with the nasal rimes, the " in the table indicates that the rime is pronounced identically to the rime above it in modern Burmese, so spelling should be memorized by lexical item.

As with in its corresponding nasal rime, learners should note that the /a/ in the /aʔ/ rime can get fronted to result in [æʔ] for many speakers.

It is important to mention that the Burmese consonant voicing phenomenon described earlier will not occur after a checked rime, so this is a good rule for learners to remember as it can prevent them from accidentally applying the consonant voicing where it is not phonotactically possible.

Non-standard Rime Spellings

With loanwords, particularly words derived from Indic, there are many other sequences spelled with coda letters that are different from the standard nasal and checked rime spellings listed above, but these will all be pronounced as one of the standard rimes described in the previous section. There are some principles that determine which spelling results in which rime pronunciation.

First, determine if the coda letter is a nasal rime coda letter or a checked rime coda letter. The nasal letters are င်, ဉ်, ဏ်, န်, မ်, same as used in the standard nasal rimes.

The checked rime coda letters are the letters associated with the standard checked rime letters, which make up the majority of the consonant initials. They can be grouped according to which of the standard checked rime coda letters they share a series with:

Note that သ် (ancestral s, modern thh) aligns with the -t coda class.

Other coda letters that do not fall into either of these two categories form a miscellaneous coda class that for the most part behaves like the nasal coda class. This class includes ယ် y, ရ် r, လ် l, ဝ် w, ဟ် h, ဠ် ḷ.

Then, examine the vowel structure on the preceding consonant. The rime output with nasal and checked coda letters is as follows:

As you can see, ော◌် with the velar series as coda simply results in the normal rimes as one would expect in standard ောင် <-aon> and ောက် <-aot>, but with coda letters from any other series (palatal, retroflex dental, labial), the rime will be <-un> (nasal) or <-ut> (checked) respectively.

In the case of a consonant with no vowel marking plus a subsequent coda letter, the velar and palatal series will result in the normal expected rimes of ◌င် <in> / ◌က် <et> and ◌ဉ် <in> / ◌စ် <it> respectively. For anything else, the result will be rimes <an> and <at>.

The one case where miscellaneous coda letters do not behave like nasal codas is in the pattern ို်, in which they become silent (completely unpronounced) instead, leaving a simple rime <o>.

For modern loanwords, if a final consonant of a syllable is pronounced, it can be written out in Burmese by using a standard checked rime coda combined with an additional coda consonant indicating the consonant sound. For example, the English word 'bus' was loaned into Burmese as ဘတ်စ်ကား ("bus car"), in which the first syllable can be pronounced as <bas> by educated speakers who are adapted to pronouncing the final -s coda. The တ် coda is still used to indicate that the standard rime coda is bat, and then the စ် makes the consonant -s. Without using the two codas together, a spelling ဘစ် would instead lead the reader to infer the pronunciation to be <bit>.

ၒ, ၓ, ၔ, ၕ
For Sanskrit transliteration, there are special letters ၒ, ၓ, ၔ, ၕ to transcribe the Sanskrit syllabic liquid vowels /r̩, r̩ː, l̩, l̩ː/ <r̥, r̥̄, l̥, l̥̄>. Their combining vowel forms are ၖ, ၗ, ၘ, ၙ. These are very rarely used.

Other Glyph Combination Irregularities

The letter လ l exhibits some irregularities in pronunciation when taking on medial glide -y-, where in some words it merged into the glide y-, and in some cases the l- and -y- are treated as broken up and pronounced separately as lăy-.

The predictable behavior can be seen when combines with the h- devoicing marker, so လျှ can indicate either initial <sh> (analogous to ယှ​) or it can be treated as broken up and pronounced separately as hlăy-.

As the pronunciation is unpredictable for လျ and လျှ, it needs to be learned word by word.

When သ <thh> takes on the h- devoicing marker as well as medial glide -y-, it becomes a rare irregular cluster သျှ that is pronounced <sh>.

Spelling Mismatches

While the rules and principles described thus far in this introduction will cover the vast majority of what a learner needs to know in order to transcribe and pronounce Burmese, there are a number of cases where the pronunciation will idiosyncratically deviate from the possible expected pronunciations, meaning that the learner will need to simply memorize the word's spelling along with its irregular pronunciation.

Some categories of these mismatches include:
Unexpected voicing of a voiceless letter
Unexpected aspiration of a voiceless letter
Unexpected aspiration of a voiced letter (only found with ဘ)
Unexpected h- devoicing ှ on a nasal letter
Unexpected drop of written medial ျ -y-
Unexpected pronunciation of medial ျ -y-
Unexpected pronunciation of written ေ as ီ
Shift to high tone before a stacked consonant sequence
Treating a single consonant as a geminate (stacked) sequence of itself (predictable before the rare လှ hl, ဋ ṭ, and ဝှ hw in Indic-derived words, must be noticed and memorized in other words)

... and many other idiosyncratically irregular mismatches that do not form overarching trends that a learner could anticipate.

Special Literary Abbreviations

There are four abbreviated syllables with special glyphs that are used for writing literary Burmese; these abbreviations are the default orthographic representation of their respective words, and the full spelling of the words is practically never used, but they are provided in the table for reference.

၏ is the possessive particle that attaches to a noun, equivalent to colloquial Burmese ရဲ့ <yę̄>. It can also be used attached to the end of a verb, functionally similar to colloquial Burmese တယ် <tę>.

၍ is a verb suffix that connects two verbs or verb clauses together.

၌ is a locative marker similar to the colloquial Burmese မှာ <hma>.

၎င်း means 'ditto, as well as.' The first syllable of this word can also get destressed into a minor syllable, resulting in lăgâon.

This glyph has an interesting origin. In the Old Burmese pronunciation of this word, the first syllable is လေး <lê>, which is identical to the Burmese number 4 (Burmese numeral: ၄), so people started to abbreviate this word with the numeral ၄. Eventually, the numeral was slightly modified to become the ၎ used in modern ၎င်း.

Numerals

The Burmese numerals are as follows:

Punctuation

၊
A small section marker that roughly corresponds to the comma in Roman script.

။
A section marker that roughly corresponds to the period (full stop) in Roman script.

Khmer is written in the Khmer script, which is ultimately derived from the Indic Brahmi script through the southern branch. As such, it is useful for learners to learn some fundamental principles underlying the typical Indic script system.

Consonants

Consonants in the Khmer script are organized into two series, the Ą Series and the Ǫ Series. The name refers to the vowel that the consonant takes as its inherent vowel, with Ą Series consonants taking /ɑː/ ( ~ /ɒː/) by default Ǫ Series consonants taking /ɔː/ by default.

Unlike Sanskrit and Pali, Khmer does not make a phonetic distinction between retroflex and dental consonants.

The two glottalized implosives /ɗ/ and /ɓ/ can vary with normal voice plosives /d/ and /b/.

វ can vary between /ʋ/ and /w/.

ឞ* and ស* are rare, used for transcribing Indic words.

The apostrophe used in Lexx Rom for glottal initial អ can be omitted at the beginning of the word, but should be used if within a word.

From the above, one can see that generally speaking, ancestrally voiceless initials induced Ą Series vowels, and ancestrally voiceless initials induced Ǫ Series vowels. However, there are exceptions, as Khmer reutilized some duplicate letters to serve as the alternate vowel series:
The unaspirated retroflex plosives ដ and ឌ are repurposed for /ɗ/ Ą Series and Ǫ Series.
The retroflex nasal ណ is repurposed for /n/ Ą Series to complement ន Ǫ Series.
The retroflex lateral ឡ is repurposed for /l/ Ą Series to complement ល Ǫ Series.
ប is used as a base letter for /ɓ/ in the Ą Series.

Even with the repurposing of these letters, there still remains some gaps where a consonant sound is missing one of the two complementary series. To meet this need, there are two superscript letters in the Khmer script ៉ and ៊ that can go on top of a base consonant letter to switch its vowel series to the other series. ៉ switches the vowel series to Ą Series, and ៊ switches the vowel series to Ǫ Series.

Series-Shifting Consonant Letters

Note that the letter ប can take both ៉ and ៊, due to it originally being an ancestral p in Indic as well as being repurposed in its base form to cover glottalized /ɓ/.

Some sonorant letters can also take an h in front of them to switch their vowel series from Ǫ Series to Ą Series, as an etymological spelling.

There are also some other supplemental consonant letters that utilize an <h> letter in front of another letter to transcribe sounds that entered Khmer through modern loanwords.

Supplemental Consonant Letters

Note that the usage of ហ្វ for transcribing /f/ means that this glyph does double duty, being used for /ʋɑː/ originally, but in modern loanwords (from French, English, etc.) indicating /fɑː/ instead.

Rimes

For every rime diacritic, learners must remember the vowel separately for the Ą Series consonants and the Ǫ Series consonants (or when there is no difference). As mentioned above, with no rime diacritic, the default rime for the Ą Series consonants is /ɑː/, and the default rime for the Ǫ Series consonants is /ɔː/.

Basic Rimes

Khmer is very particular with its vowel distinctions, so learners should take the time to practice them with native speakers.

Rimes with Coda

Typically, vowels with glottal stop/null initial consonant are attached to អ, but there are also independent vowel letters that are used mainly in spelling Indic words. The exact rime used for these letters can vary by word, so learners must learn them with each vocabulary item.

Independent Vowel Letters

ឨ is no longer used in modern Khmer.

The Bouyei-Zhuang-Tay-Nung cluster of languages belong to both Northern Tai (Bouyei and northern Zhuang) and Central Tai (southern Zhuang and Tay-Nung) branches. These languages are traditionally written in Sinographs (漢字), also known as Chinese characters, but nowadays romanization systems are the officially sanctioned writing systems, and use of Chinese characters has decreased, mainly being limited to informal usage. Chinese character usage has never been officially standardized in these languages, so usage of characters can vary from area to area and writer to writer, though in general, character usage seems to have been more consistent within Tay than the Zhuang varieties.

The romanization system for Standard Zhuang recognized in the PRC is a typical PRC Pinyin style romanization scheme, which reutilizes some of the principles of the Hanyu Pinyin in use for Standard Mandarin, but with many other features that will be unfamiliar to Hanyu Pinyin users, such as use of letters appended at the end of a syllable to indicate the tone, instead of the diacritic marks such as á and à. Lexxify Hub also has a Lexx Rom system romanization to help learners get used to pronouncing Zhuang; the Lexx Rom system on Lexxify Hub is made to be extremely consistent with the systems in place for other Tai languages such as Thai, Lao, and Shan.

Initials

Typical of Northern Zhuang, there are no aspirated plosive consonants, but there is a distinction between the unaspirated /p/ and the implosive /ɓ/, as well as unaspirated /t/ and implosive /ɗ/. /w/ in the Zhuang-speaking area can often vary in realization with [β] or [v].

Rimes

Plain Rimes

/ɐ/ and /aː/ contrast in both height and length, as in Cantonese.

Nasal Coda Rimes

Checked Coda Rimes

Tones

Shan is written in an version of the Mon-Burmese Myanmar script, and its modern  orthography has been standardized to match up quite precisely with its phonology, so it is overall easier to learn how to read and write compared to Thai.

Consonants

Main Consonant Initials

Many speakers merge /r/ into /l/. /ts/ can be realized as /tɕ/. Some speakers merge /ts/ into /s/.

Additional Consonant Initials

These consonants are only sporadically used, for transcribing sounds in loanwords from other languages.

Codas

Shan has six final coda consonants (not including glides, which will be discussed in the rimes section).

Vowels and Rimes

Basic Vowels

As in Thai, // is quite low, approaching .

Compound Rimes

Tones

Tones

The sixth tone is only used for emphasis in general Shan orthography (though it is also used in the northern dialect).  

The general etymological correspondences between Shan tones and tones in other Tai languages as well as broader East Asia is as follows:

Tone 1 corresponds to A1 (Yinping).
Tone 2 corresponds to B1 (Yinqu) or DL1 (Long Yinru).
Tone 3 corresponds to C1 (Yinshang), B2 (Yangqu), and DL2 (Long Yangru).
Tone 4 corresponds to A2 (Yangping), DS1 (Short Yinru).
Tone 5 corresponds to C2 (Yangshang), DS2 (Short Yangru).

There are various dialects and accents of the Lanna language, among which that of Chiang Mai is often viewed as a prestige accent, so that is used as the basis of the Lexx Rom transcription. While Lanna is used as a regional lingua franca in northern Thailand, it is not currently in use as an official or standardized language, so there is a great amount of variation among speakers, with various levels of admixture and code-switching with standard Central Thai (the official language of Thailand) as well as other local languages.

Lanna is historically written in a version of the Tai Tham script, a script of the northern Thailand area that is also historically used for other Tai languages such as Tai Lue, Khuen, and occasionally for Lao, with some subtle differences in usage and appearance of glyphs between the different languages. The letters of the Tai Tham script are overall quite similar to the Burmese script, both being derived from the Old Mon script. This is in contrast to the (central) Thai script and Lao script, which are derived from Old Khmer script, but both Old Mon and Old Khmer scripts are ultimately derived part of the southern family of Indic Brahmic scripts. In contemporary northern Thailand, Thai script has become the dominant script used for writing Lanna, so literacy in Tai Tham script has waned, but it is still used in local monasteries and some other domains as a heritage script.

Because the Tai Tham and Thai scripts are based on the Indic script system, it is helpful to learn some of the basic principles in order to understand the underlying logic and rationale beneath its structure. Additionally, the Thai tonal system developed as part of the East Asian Sinospheric tone system Sprachbund, so it is helpful to learn some principles of how these tonal systems work in order to understand how Lanna tones interplay with the Indic-based script.

In the following chart of initial consonants, the column "Ancestral Indic" shows what sound the letter originally transcribed in Sanskrit/Pali, "Ancestral Lanna" shows what sound the letter would have originally transcribed in the past (different from modern Lanna), and "Ancestral Tone Class" shows what tone class the letter belongs to based on its original phonetic value.

Initials

The three tone classes in Lanna are High, Middle, and Low, based on the ancestral sound of each letter.

Ancestral unaspirated voiceless plosives/affricate (ᨠ/ก k, ᨧ/จ tʃ, ᨭ/ฏ ʈ, ᨲ/ต t, ᨸ/ป p), voiceless aspirated plosives/affricate (ᨡ/ข kʰ, ᨨ/ฉ tʃʰ, ᨮ/ฐ ʈʰ, ᨳ/ถ tʰ, ᨹ/ผ pʰ) and ancestral voiceless fricatives (ᨺ/ฝ f, ᩆ/ศ ʃ, ᩇ/ษ ʂ, ᩈ/ส s, ᩉ/ห h) comprise the High tone class.

Ancestral  glottalized initials (ᨯ/ด ɗ, ᨷ/บ ɓ, ᩀ/อย ʔj, ᩋ/อ ʔ) comprise the Middle tone class.

It is important to note that in contrast to Standard Thai, the unaspirated voiceless group of initials belongs to the High tone class, rather than the Middle tone class (which is only left with four initials).

Ancestral voiced plosives/affricate (ᨣ/ค ɡ, ᨩ/ช dʒ, ฑ ɖ, ᨴ/ท d, ᨻ/พ b), ancestral breathy voiced plosives/affricate (ᨥ/ฆ ɡʱ, ᨫ/ฌ dʒʱ, ᨰ/ฒ ɖʱ, ᨵ/ธ dʱ, ᨽ/ภ bʱ), ancestral voiced fricatives (ᨤ/ฅ ɣ, ᨪ/ซ z, ᨼ/ฟ v, ᩌ/ฮ ɦ), and ancestral sonorants (ᨦ/ง ŋ, ᨬ/ญ ɲ, ᨱ/ณ ɳ, ᨶ/น n, ᨾ/ม m, ᨿ/ย j, ᩁ/ร r, ᩃ/ล l, ᩅ/ว w, ᩊ/ฬ ɭ) comprise the Low tone class.

Some sonorants can also have a ᩉ/ห h appended in front of them, and become High tone class versions of that consonant (this does not affect the pronunciation of the initial consonant in modern Lanna): ᩉ᩠/หง, ᩉ᩠ᨿ/หญ, ᩉ᩠ᨶ/หน, ᩉ᩠ᨾ/หม, ᩉᩕ/หร, ᩉ᩠ᩃ/หล, ᩉ᩠ᩅ/หว.

The names High, Middle, and Low refer to areal pattern in East Asia of voiced initials triggering lower tones for their syllables compared to the syllables with voiceless initials, which is a cross-linguistically attested natural acoustic tendency due to the lower formant frequency caused by the voicing. However, this no longer reflects the modern Lanna tone reflexes for these letters, due to language change over time.

As seen in the chart above by comparing the ancestral sounds to the modern Lanna sounds, voiced plosives became unaspirated voiceless consonants, while breathy voiced plosives became aspirated voiceless consonants (in Central Thai they all became aspirated voiceless consonants). Voiced fricatives became voiceless as well. There was also some change with the fricatives that differs from Central Thai, with ɣ becoming kʰ, while tʃʰ became s.

Ancestral r became either l or a Low tone class h (as opposed to High tone class h). In general, in native Tai lexical stock it merged into h, while in loanwords such as from Sanskrit and Pali it merged into l, but there are inconsistencies and exceptions to watch out for. Also note that with modern loanwords there are now words that use pronounce the initial r again.

Some speakers pronounce /kʰ/ as [x].

Note that there is only one series of coronal consonants in Lanna; the retroflex consonant letters are used for transcribing Sanskrit/Pali loanwords, and are pronounced the same as their dental counterparts.

Unlike in Standard Thai, Lanna does not pronounce initial clusters with /l/ or /r/. /l/ in an initial cluster is dropped, while /r/ in an initial cluster triggers aspiration on a remaining unaspirated consonant (remember that standalone r has in many cases merged into h), so that for example what is written <pr> becomes pronounced as /pʰ/.

Tai Tham consonant symbols are referred to by the consonant sound plus short /a/, unlike in Thai where they are referred to the consonant sound plus /ɔː/.

Final Codas

While there are many initial consonants in Lanna, augmented even further by extra letters dedicated for transcribing Indic sounds, there are only six options for final codas (plus semivowels -j and -w, which will be shown later in the rime section below), so all the dozens of consonant letters must collocate into one of these six options, based on what sound is the most similar.

These six codas -k, -ng, -t, -n, -p, -m are a common set of final codas in the East Asia region.

Vowels and Rimes

As like in Thai and Lao, Lanna has nine basic vowels, all of which can occur as short or long. The representation of the vowel in the script is sometimes different if there is a final coda consonant coming after it.

Basic Vowels

Lanna has the same nine basic vowels as Thai and Lao, in short and long forms. As in Thai and Lao, short vowels without an ending coda consonant afterwards actually have the glottal stop coda, and count as checked coda syllables for the purposes of tonal behavior (discussed in tone section below).

As in Burmese orthography, the tall aa (ᩤ) is used specifically for certain consonant symbols where the default semicircle aa (ᩣ) could result in a shape that would have looked too similar to a different consonant letter. The symbols that take the tall aa are:

ᨣ (to distinguish from ᨠ)
ᨴ (to distinguish from ᨽ)
ᨵ (to distinguish from ᨲ)
ᨷ (to distinguish from ᩉ)
ᩅ (to distinguish from ᨲ)

Falling Diphthongs

The standalone falling diphthongs are very marginal to the phonological system, so rather than <ia, iia> etc. which would in practice mean that mostly <iia> is encountered, Lexx Rom indicates these as <iah, ia> etc.

Other Rimes

The ไ◌, ใ◌ originally contrasted, with ใ◌ indicating /aɰ/, though these have now merged in Lanna. The sounds are still kept distinct in some other Tai languages such as Shan.

Tones

Lanna tones can vary from area to area, so the description below applies to Chiang Mai Lanna. Chiang Mai Lanna tones are for the most part quite similar to the tones in Bangkok Thai, with just one more extra tone, making for a total of six tones. In the numbering system below, the first three tones are the main tones for the High tone class, and the next three tones are the main tones for the Low tone class.

Tones

The description for tones in Chiang Mai Lanna can vary depending on the focus of the author.

Tone 1 ǎa is similar to Thai ǎa, being a low rising tone that can have some dipping at the beginning before the rising.

Tone 2 a̩a is similar to Thai a̩a, being the cross-linguistically common bottom of vocal range tone.

Tone 3 is a high tone that can carry a little bit of creaky glottalization towards the end. It can also have a little bit of downdrift in pitch at the end as well, so it has been described by some as a high falling tone ˥˥˦ also. In general though, it is easiest for learners to conceptualize this tone as a high flat tone when learning.

Tone 4 is a mid-flat tone that can have some downdrift towards the end, similar to the Thai equivalent.

Tone 5 is a high falling tone that can have a little bit of peaking as in Thai. Some manuals may list Tone 3 as a high falling tone and then list this tone as a falling tone, so we prefer to characterize Tone 3 as a high flat tone with some downdrift and creakiness as opposed to this tone which is a more typical high falling tone.

Tone 6 is a high rising tone, similar to the high rising tone in Thai (which is often just called the high tone in Thai).  

As in Thai and Lao orthography, tones are not marked out directly in Lanna. The tone for a syllable is inferred from the initial consonant's tone class, presence of any tone mark, and syllable type (non-checked vs. short checked vs. long checked).

As mentioned previously, there are three tone classes in Lanna (High tone class, Middle tone class, and Low tone class), and the tone of a syllable will depend on which tone class the initial belongs to. The easiest way to conceptualize how to calculate Chiang Mai Lanna tones is by remembering patterns for the High tone class and the Low tone class, and then the Middle tone class mostly falls in line with the High tone class except in one case where it aligns with the Low tone class instead.

Remember that the Middle Tone Class in Lanna only consists of four initials (ᨯ/ด ɗ, ᨷ/บ ɓ, ᩀ/อย ʔj, ᩋ/อ ʔ), which is less than in Thai or Lao. The ancestral unaspirated voiceless plosives/affricate (ᨠ/ก k, ᨧ/จ tʃ, ᨭ/ฏ ʈ, ᨲ/ต t, ᨸ/ป p) instead merged into High tone class.

In the High Tone Class:

The default tone category (no tone mark) is low rising tone <ǎa>.
With tone mark one (◌່ ), the tone becomes low tone <a̩a>.
With tone mark two (◌້ ), the tone becomes high creaky tone <āa>.

If the syllable has a checked coda, one must check whether it is a short or long rime.
For short rime checked coda syllables, the tone becomes low rising tone <ǎa>.
For long rime checked coda syllables, the tone becomes low tone <a̩a>.

In the Low Tone Class:

The default tone category (no tone mark) is mid-flat tone <aa>.
With tone mark one (◌່ ), the tone becomes high falling tone <âa>.
With tone mark two (◌້ ), the tone becomes  high rising tone <áa>.

If the syllable has a checked coda, one must check whether it is a short or long rime.
For short rime checked coda syllables, the tone becomes high rising tone <á>.
For long rime checked coda syllables, the tone becomes high falling <âa>.

In the Middle Tone Class:

In Chiang Mai Lanna, all tonal behavior for the Middle tone class aligns with the High tone class except in the default tone category (no tone mark), in which case it aligns with the Low tone class instead (mid-flat tone <aa>).

Besides the tonal behavior listed above for the three tone categories, one may sometimes see the case where a word with a consonant in the High tone class carries the high rising tone <áa>. In the Thai script, this can be accommodated with the use of tone mark three (◌๊ ), which overtly marks the high rising tone onto a syllable. In the Tai Tham script for Lanna there is traditionally no tone mark three available, but the Tai Tham orthography for the Khuen language does have a tone mark three ( ᩷ ) so this will be recommended on Lexxify Hub.

The Thai script also has a tone mark four (◌๋ ) that overtly indicates a low rising tone. Theoretically there is even less utility for a tone mark four than in Lanna, but in the case of onomatopoeia or neologisms/loanwords, the Khuen tone mark four (  ᩸ ) can be used.

The general etymological correspondences between Lanna tones and tones in other Tai languages as well as broader East Asia is as follows:

Tone 1 corresponds to A1 (Yinping) and DS1 (Short Yinru).
Tone 2 corresponds to B1 (Yinqu) and DL1 (Long Yinru).
Tone 3 corresponds to C1 (Yinshang).
Tone 4 corresponds to A2 (Yangping).
Tone 5 corresponds to B2 (Yangqu) and DL2 (Long Yangru).
Tone 6 corresponds to C2 (Yangshang) and DS2 (Short Yangru).

Other notes on the Tai Tham Lanna writing system

ᩢ is /a/ when there is a coda present, but in some words it also does double duty as a -k coda.

ᩴ can indicate a -m or -ng coda, and in some words it can indicate a vowel /ɔː/.

ᨶᩣ is a special ligature for the sequence /naː/.

ᩓ is a special ligature for the sequence /lɛː/.

Like its Burmese equivalent, ᩔ represents a geminate ss in words from Sanskrit/Pali.

There are a couple varieties of Lao-Isan, of which the Vientiane dialect is the most commonly viewed as a prestige accent, so the Lexx Rom tones will be based on this pronunciation.

The Lao script is based on the Old Khmer script, which is based on the Indic Brahmic script model that is widespread across South and Southeast Asia. The shapes of the letters are very similar to their corresponding letters in Thai script. Therefore, when learning Lao script as a non-native, even if never having studied an Indic script before, it is helpful to learn some of the basic principles in order to understand the underlying logic and rationale beneath its structure. Additionally, the Lao tonal system developed as part of the East Asian Sinospheric tone system Sprachbund, so it is helpful to learn some principles of how these tonal systems work in order to understand how Lao tone marks interplay with the Indic-based script.

In the following chart of initial consonants, "Ancestral Lao" shows what sound the letter would have represented in earlier forms of the language, and "Ancestral Tone Class" shows what tone class the letter belongs to based on its original phonetic value.

Compared to Thai orthography, modern Lao orthography has underwent a massive simplification, so a lot of letters that Thai still uses

Initials

Some dialects of Lao spirantize /kʰ/ into /x/.

/r/ has merged into /l/ or /h/ in modern Lao; words that merged to /h/ are spelled with ຮ, and words that merged to /l/ are spelled with ລ. However, there are also modern loanwords in Lao that have /r/, and so this letter is used for the /r/ sound for in these modern loanwords. There are also some writers who prefer to etymologically use ຣ to indicate the /l/ in words where it was etymologically originally /r/ such as in Sanskrit/Pali loanwords, but this is not officially sanctioned in the modern orthography.

The pronunciation of ວ can vary between /ʋ/ and /w/, and although there is a tendency towards /ʋ/ in the speech of many from Vientiane, for consistency with other Lao dialects as well as other Tai languages, it is transcribed as <w> in Lexx Rom.

The three tone classes in Thai are High, Middle, and Low, based on the ancestral sound of each letter.

Voiceless aspirated plosives/affricate (ຂ kʰ, ຖ tʰ, ຜ pʰ) and voiceless fricatives (ສ s, ຝ f, ຫ h) comprise the High tone class.

Unaspirated voiceless plosives/affricate (ກ k, ຈ tʃ, ຕ t, ປ p) and glottalized initials (ດ ɗ, ບ ɓ, ຢ ʔj, ອ ʔ) comprise the Middle tone class.

Voiced plosives/affricate (ຄ ɡ, ຊ dʒ, ທ d, ພ b), voiced fricatives (ซ z, ຟ v, ຮ ɦ), and sonorants (ງ ŋ, ຍ ɲ, ณ ɳ, ນ n, ມ m, ຣ r, ລ l, ວ w) comprise the Low tone class.

ຢ is a special case because it is a Middle tone class letter despite being a sonorant /j/. This is because technically this letter corresponds to ancestral glottalized /ˀj/, whereas the true ancestral voiced /j/ merged into ຍ ɲ. When the -j coda is written as part of a compound rime, the ຍ ɲ letter is used, rather than ຢ (detailed in the rime section later).

The sonorant letters can also become High tone class if they have the letter ຫ h appended in front of them, as ຫງ ຫຍ ຫນ ຫມ ຫລ ຫວ. There are also some special ligature characters for some of these than can be used: ໜ (ຫນ), ໝ (ຫມ), and ຫຼ (ຫລ).

The names High, Middle, and Low refer to areal pattern in East Asia of voiced initials triggering lower tones for their syllables compared to the syllables with voiceless initials, which is a cross-linguistically attested natural acoustic tendency due to the lower formant frequency caused by the voicing. However, this no longer reflects the modern Lao tone reflexes for these letters, due to language change over time.

As seen in the chart above by comparing the ancestral sounds to the modern Lao sounds, voiced plosives/fricatives devoiced and became aspirated.

Final Codas

These six codas -k, -ng, -t, -n, -p, -m are a common set of final codas in the East Asia region. Note that the letter used as the default to write finals -t and -p are actually the letters d and b.

Vowels and Rimes

Lao has nine basic vowels, all of which can occur as short or long. The representation of the vowel in the script is sometimes different if there is a final coda consonant coming after it.

Basic Vowels

/ɛ(ː)/ is quite low, close to /æ(ː)/. Short /a/ can approach /ɐ/, though not quite exactly reaching the position of Cantonese /ɐ/.

Short vowels without any coda consonant ending actually carry a light glottal stop coda /ʔ/, though it is not always fully pronounced if not in word-final position, and is sometimes omitted in rapid speech as well. This glottal coda means that these short vowel rimes actually count as checked syllables for the purposes of tonal behavior (discussed in the next section). This glottal stop coda is not indicated in Lexx Rom transcription.

Compound Rimes

The short falling diphthongs /iə, uə, ɯə/ are marginal to the phonological system, so the default Lexx Rom <ia, ua, ưa> indicate the full length diphthongs /iːə, uːə, ɯːə/.

The letter ຽ was formerly also used as an alternate letter for ຍ in any non-intial position, but this is not part of the modern official orthography.

The mark ໃ◌ indicates a different sound /aɰ/, still found in other Lao dialects as well as Tai languages such as Shan, but in Vientiane Lao, this sound has merged into /aj/, though the spelling still reflects it etymologically. Likewise, a few rare words with /aj/ spelled ◌ັຍ (generally due to etymological transliteration of Indic loans) must have the spelling memorized.

Tones

Vientiane Lao has five contrastive tones, though the script does not directly indicate them in a one-to-one corresponding manner. Tones in Lao are indicated through an intricate interaction between initial consonant tone class (refer to the initials section of this guide) and tone mark. The five tones are as follows:

Tones

The commonly used tone names reflect a naming convention in Thai and Lao that follows the general tone contour pattern, but it needs to be pointed out that technically the high tone is also a rising tone contour, and the low tone is also a falling tone contour, so you may sometimes see materials use wordier names such as "high falling" vs. "low falling" vs. "high rising" vs. "low rising", and this is also complicated by the differences between tone contours across different Lao dialects as well as free variation between different speakers, for example the <âa> is technically a peaking tone ˦˥˧ for many speakers, though it has also been described as a high falling tone, and the (low) rising tone can actually have a dipping contour before rising.

As mentioned previously, there are three tone classes in Lao, High tone class, Middle tone class, and Low tone class, and the tone of a syllable will depend on which tone class the initial belongs to. The easiest way to conceptualize how to calculate Vientiane Lao tones is by remembering patterns for the High tone class and the Low tone class, because in Vientiane Lao the tone for the Middle tone class will always pattern with either the High or Low tone class, rather than having its own set of tones.

In the High Tone Class:

The default tone category (no tone mark) is <ǎa>.
With tone mark one (◌່ ), the tone becomes mid-flat tone <aa>.
With tone mark two (◌້ ), the tone becomes low tone <a̩a>.

If the syllable has a checked coda, one must check whether it is a short or long rime.
For short rime checked coda syllables, the tone becomes high rising <á>.
For long rime checked coda syllables, the tone becomes low tone <a̩a>.

In the Low Tone Class:

The default tone category (no tone mark) is <áa>.
With tone mark one (◌່ ), the tone becomes mid-flat tone <aa> (same as in the High Tone Class).
With tone mark two (◌້ ), the tone becomes high falling tone <âa>.

If the syllable has a checked coda, one must check whether it is a short or long rime.
For short rime checked coda syllables, the tone becomes mid-flat tone <a>.
For long rime checked coda syllables, the tone becomes high falling <âa>.

In the Middle Tone Class:

In Vientiane Lao, all tonal behavior for the Middle tone class will align with either of the High or Low tone classes, so learners can just memorize which category causes it to align with either High or Low.

The default tone category (no tone mark) is <ǎa> ( = High tone class).
With tone mark one (◌່ ), the tone becomes mid-flat tone <aa> ( = High & Low tone classes).
With tone mark two (◌້ ), the tone becomes high falling tone <âa> ( = Low tone class).

If the syllable has a checked coda, one must check whether it is a short or long rime.
For short rime checked coda syllables, the tone becomes mid-rising tone <á> ( = High tone class).
For long rime checked coda syllables, the tone becomes high falling <âa> ( = Low tone class).

The only consistent tone mark across all Vientiane Lao syllables is tone mark one (◌່ ), which indicates mid-flat tone regardless of tone class.

There are also two additional tone marks, tone mark 3 (◌໊ ) and tone mark 4 (◌໋​ ), which are very rarely used in Lao. If seen, these overtly indicate high rising tone <áa> and low rising tone <ǎa> respectively, as they do in Thai.

As mentioned earlier, there are many different dialects and accents of Lao, and even within Vientiane Lao, you will encounter variation between speakers, so for example the default tone for Middle tone class for some speakers is a sixth tone- a low flat tone /˩/ (a̠a). But in general, the tone values as given above are a good reference point for learners of Vientiane Lao.

The general etymological correspondences between Lao tones and tones in other Tai languages as well as broader East Asia is as follows:

Tone 1 corresponds to B (Qu) and DS2 (Short Yangru).
Tone 2 corresponds to C1 (Yinshang) and DS2 (Long Yinru).
Tone 3 corresponds to C2 (Yangshang) and DL2 (Long Yangru).
Tone 4 corresponds to A2 (Yangping) and DS1 (Short Yinru).
Tone 5 corresponds to A1 (Yinping).

The Thai script is based on the Old Khmer script, which is based on the Indic Brahmic script model that is widespread across South and Southeast Asia. Therefore, when learning Thai script as a non-native, even if never having studied an Indic script before, it is helpful to learn some of the basic principles in order to understand the underlying logic and rationale beneath its structure. Additionally, the Thai tonal system developed as part of the East Asian Sinospheric tone system Sprachbund, so it is helpful to learn some principles of how these tonal systems work in order to understand how Thai tone marks interplay with the Indic-based script.

In the following chart of initial consonants, the column "Ancestral Indic" shows what sound the letter originally transcribed in Sanskrit/Pali, "Ancestral Thai" shows what sound the letter would have originally transcribed in the past (different from modern Thai), and "Ancestral Tone Class" shows what tone class the letter belongs to based on its original phonetic value. The letters are named by the initial sound plus -ǫǫ for Middle and Low tone class or -ǫ̌ǫ for High tone class, followed by a word that uses the letter.

Initials

As in many Southeast Asian Indic orthographies, the Indic short a letter is used in Thai as the null/glottal stop initial. In Lexx Rom, this sound can be omitted from transcription at the beginning of a word.

The three tone classes in Thai are High, Middle, and Low, based on the ancestral sound of each letter. The names High, Middle, and Low refer to areal pattern in East Asia of voiced initials triggering lower tones for their syllables compared to the syllables with voiceless initials, which is a cross-linguistically attested natural acoustic tendency due to the lower formant frequency caused by the voicing. However, this no longer reflects the modern Thai tone reflexes for these letters, due to language change over time.

Voiceless aspirated plosives/affricate (ข kh, ฉ ch, ฐ ṭh, ถ th, ผ ph) and voiceless fricatives (ฃ x, ฝ f, ศ ś, ษ ṣ, ส s, ห h) comprise the High tone class.

Unaspirated voiceless plosives/affricate (ก k, จ c, ฏ ṭ, ต t, ป p) and glottalized initials (ฎ đ̣, ด đ, บ ƀ, อ ') comprise the Middle tone class.

Voiced plosives/affricate (ค g, ช j, ฑ ḍ, ท d, พ b), breathy voiced initials (ฆ gh, ฌ jh, ฒ ḍh, ธ dh, ภ bh, ฮ h̤), voiced fricatives (ฅ ğ, ซ z, ฟ v), and sonorants (ง ṅ, ญ ñ, ณ ṇ, น n, ม m, ย y, ร r, ล l, ว w, ฬ ḷ) comprise the Low tone class.

Some sonorants can also have a ห h appended in front of them, and become High tone class versions of that consonant (this does not affect the pronunciation of the initial consonant in modern Thai): หง hṅ, หญ hñ, หน hm, หม hn, หย hy, หร hr, หล hl, หว hw. อ (') can also get appended in front of ย y in a couple words, changing it to a Middle tone consonant.

As seen in the chart above by comparing the ancestral sounds to the modern Thai sounds, voiced plosives/fricatives and breathy voiced initials devoiced and became aspirated.

Thai no longer has velar fricatives, so ฃ x and ฅ ğ are now obsolete, pronounced like ข kh and ค g respectively (modern kʰ in High and Low tone classes).

/r/ is generally merged with /l/ in casual Thai, but it is preserved in formal speech, and Thai people generally place great importance on using it when teaching foreigners. In casual Thai, as the second component of an initial cluster, it is commonly dropped.

Note that there is only one series of coronal consonants in Thai; the retroflex consonant letters are used for transcribing Sanskrit/Pali loanwords, and are pronounced the same as their dental counterparts.

As modern Thai does not have a voiced /ɡ/ or /dʒ/ initial, these sounds in modern loanwords are transcribed using the letters for the corresponding voiceless sounds ก k and จ c, and in casual spelling in Roman script they are often spelled as g and j, something followed in some Romanization systems for convenience.

Final Codas

While there are many initial consonants in Thai, augmented even further by extra letters dedicated for transcribing Indic sounds, there are only six options for final codas (plus semivowels -j and -w, which will be shown later in the rime section below), so all the dozens of consonant letters must collocate into one of these six options, based on what sound is the most similar.

These six codas -k, -ng, -t, -n, -p, -m are a common set of final codas in the East Asia region. Note that the letter used as the default to write finals -t and -p are actually the letters for initial /d/ and /b/.

Vowels and Rimes

Thai has nine basic vowels, all of which can occur as short or long. The representation of the vowel in the script is sometimes different if there is a final coda consonant coming after it. The names of these letters are composed of the word สระ sa̩ra̩ + the vowel sound. The null/glottal stop initial is used as the base carrier for these vowels.

Basic Vowels

/ɛ(ː)/ is quite low, close to /æ(ː)/. Short /a/ can approach /ɐ/, though not quite exactly reaching the position of Cantonese /ɐ/.

Short vowels without any coda consonant ending actually carry a light glottal stop coda /ʔ/, though it is not always fully pronounced if not in word-final position, and is sometimes omitted in rapid speech as well. This glottal coda means that these short vowel rimes actually count as checked syllables for the purposes of tonal behavior (discussed in the next section). This glottal stop coda is not indicated in Lexx Rom transcription.

Note that the null/glottal stop initial อ also does double duty as a potential vowel indicator in some vowel rimes , so for example ออ indicates <ǫǫ>.

Falling Diphthongs

The short falling diphthongs /iə, uə, ɯə/ are marginal to the phonological system and do not occur with a coda letter following, so the default Lexx Rom <ia, ua, ưa> indicate the full length diphthongs /iːə, uːə, ɯːə/.

Other Compound Rimes

The mark ใ◌ once indicated a different sound /aɰ/, still found in other Tai languages such as Shan, but in Thai, this sound has merged into /aj/, though the spelling still reflects it in a handful of high-frequency words for which the spelling must be memorized. Likewise, words with /aj/ spelled as ไ◌ย, ◌ัย (generally due to etymological transliteration of Indic loans) must have the spelling memorized.

Special Rimes

ฤ, ฤๅ, ฦ, ฦๅ are letters for the Sanskrit syllabic liquids /r̩, r̩ː, l̩, l̩ː/ <r̥, r̥̄, l̥, l̥̄>, which are relatively uncommon. In some words, the inherent vowel for these letters will be /i/ instead of /ɯ/, which must be memorized word by word.

Tones

Thai has five contrastive tones, though the script does not directly indicate them in a one-to-one corresponding manner. Tones in Thai are indicated through an intricate interaction between initial consonant tone class (refer to the initials section of this guide) and tone mark. Syllables in the Middle tone class that do not have a checked coda (-p, -t, -k, or -ʔ) will be used as reference, because the tones with tone marks correspond neatly as follows:

Tones (ordered by Middle tone class)

Tone 1 is a mid-flat tone that can have some downdrift towards the end.
Tone 2 is a tone that is at the bottom of the vocal register, which cross-linguistically is ambiguous between being a low level tone or a low falling tone.
Tone 3 is a high falling tone that often carries some peaking during the higher segment, so it is often described as a peaking tone as well.
Tone 4 is a high tone that rises higher.
Tone 5 is a low rising tone that can have some dipping at the beginning, and in rapid speech in non-final position it can also surface as a low falling tone (without having time to rise up again).

High tone class:
For syllables in the High tone class that do not have a checked coda, the default tone if there is no tone mark is instead <ǎa>.
Tone marks ◌​่ and ◌้ will result in the same tone as in the above chart for Middle tone class (<a̩a> and <âa>).
Tone marks ◌๊ and ◌๋ are not used with the High tone class.

Low tone class:
For syllables in the Low tone class that do not have a checked coda, the default tone if there is no tone mark is the same as in the Middle tone class (<aa>).
However, the tones when combined with tone marks ◌​่ and ◌้ are <âa> and <áa> respectively, which differs from the pattern seen in the Middle/High tone classes. Tone marks ◌๊ and ◌๋  are not used with the Low tone class.

In checked syllables:
For syllables that have a checked coda (-p, -t, -k, or -ʔ), if the initial is Middle or High tone class, the default tone (no tone mark) will be <a̩>.
For syllables with initial in the Low tone class with default tone (no tone mark), one must check another step further and determine whether it is a short vowel rime or a long vowel rime; short vowel rimes will take tone <á>, and long vowel rimes will take tone <âa>. The short <á> in these checked syllables may not have that much time to rise up in tone, so it can surface as a [˦] in rapid speech.

Syllables with checked codas will rarely take tone marks; these only occur in particles, onomatopoeia, or loanwords. If taking a tone mark, the tone becomes the equivalent of whatever tone the syllable would have been if there had not been a checked coda present.

In addition to these tones, there is also an neutralized tone that occurs only as a short a in the first segment of a sesquisyllabic word. This is not indicated with any tone mark in Lexx Rom (recall that the short <a> would otherwise be <a>̩ or <á>, depending on tone class of the initial, so there is no ambiguity).

The general etymological correspondences between Thai tones and tones in other Tai languages as well as in broader East Asia is as follows:

Tone 1 corresponds to A2 (Yangping).
Tone 2 corresponds to B1 (Yinqu) and D1 (Yinru).
Tone 3 corresponds to C1 (Yinshang), B2 (Yangqu) and DL2 (Long Yangru).
Tone 4 corresponds to C2 (Yangshang) and DS2 (Short Yangru).
Tone 5 corresponds to A1 (Yinping).

Other Symbols

◌์ on top of a consonant letter cancels its entire sound out, meaning the entire letter is silent, and only retained for etymological purposes.

ๆ reduplicates the prior word.

ฯ marks an abbreviation.

◌​ฺ under a consonant letter can be used in Indic transliteration to indicate that the inherent vowel has been canceled out (but not the entire letter as in ◌์). This mark is not used in written Thai otherwise, so it is not important for beginners.

Numerals

The mainstream prestige accent of Taiwanese is that of Tainan, the old capital of Taiwan, and that is the reference accent for Taiwanese (TW) on Lexxify Hub. The Tainan accent is tilted towards Zhangzhou in rime isoglosses, and compared to the Tainan accent, the Taipei accent is tilted more towards Xiamen, though neither is exactly Zhangzhou nor Xiamen. In terms of intelligibility with Hokkien, itelligibility between Taiwanese and Xiamen Hokkien is quite good, but as one goes more towards the extremes ends of the Zhangzhou-Quanzhou spectrum, intelligibility will tend to decrease, and intelligibility is also lower with the inland lects of Longyan and Datian as well as with Chaoshanese. Taiwanese in general has a higher preference for utilizing literary readings of characters in compound words where many Hokkien dialects and also Chaoshanese will use the plain readings of characters.

Initials

Initials

As in Standard Mandarin, the initials /tɕ/, /tɕʰ/, /dʑ/, /ɕ/ only occur before the high front vowel (in Taiwanese, only /i/ due to the lack of /y/), so they can be considered allophones of /ts/, /tsʰ/, /dz/, /s/ respectively when in front of /i/. This is reflected in the Peh-oe-ji and Tai-lo romanization systems, which use the same symbols for both sets of initials. The Zhuyin system as well as Hanyu Pinyin overtly indicate this difference, so as Lexx Rom for Sinitic languages is an expanded Hanyu Pinyin system, it will also reflect this difference. This palatalization is obligatory in Taiwanese, whereas in Hokkien dialects of southern Fujian it can vary.

The voiced /dz/, /dʑ/ initials are inherited from Zhangzhou Hokkien, but in Xiamen and Quanzhou, it merged into /l/. This merger is also quite strong in Taiwan, being increasingly dominant, even in the south, so if learners are having difficulty with this sound, it is very much so acceptable to replace it with the /l/ sound.

The initials /b/ and /ɡ/ may carry some optional pre-nasalization becoming realized as [ᵐb] and [ᵑɡ].

In Taiwanese as well as many dialects of Hokkien, the nasal phonemes /m/, /n/, and /ŋ/ can actually be considered allophones of /b/, /l/, and /ɡ/ respectively, when in front of a nasalized vowel rime (not nasal coda rimes). This is different from the phonological situation in Chaoshanese. For ease of romanization and for the sake of simplicity for learners, they are indicated overtly in the vast majority of transcription systems, including the Zhuyin, Peh-oe-ji, Tai-lo, and Lexx Rom used on Lexxify Hub. The nasalization of the vowel is subsequently not indicated, as it is not contrasted against a non-nasalized vowel in this environment. Therefore, the plain reading of 滿 /muã˥˧/ is, for example, transcribed as moá in Peh-oe-ji instead of boáⁿ (which works phonologically, but due to being more opaque would require more time for learners to internalize the phonological process).

Rimes

Plain Rimes

Mainstream Taiwanese has six basic vowel phonemes, /i/, /u/, /a/, /e/, /ə ~ ɤ/, and /ɔ/. Of these, the /ə/ phoneme is historically realized as a /o/ reflected in Peh-oe-ji and Tai-lo, and it is still realized as such in parts of northern Taiwan (sometimes with weaker rounding), but from around Taichung southwards the /ə/ pronunciation is dominant, and this is reflected explicitly in the contemporary Zhuyin with the character <ㄜ>, as well as in Lexx Rom with <eu>. If a learner is having difficulty producing a less-rounded /o̜/, it is preferable to aim for producing an /ə/ rather than going for a diphthongized /ou/. Some speakers, particularly younger generations, will also merge this relatively unstable /o/ into the rime /ɔ/. To further complicate matters, /ə/ can also appear as a separate vowel from both /o/ and /ɔ/ in Lukang Quanzhou-style accent (as in Quanzhou Hokkien), though this is seldomly encountered day-to-day in Taiwan and not applicable to mainstream Taiwanese.

Nasalized Rimes

In the mainstream Taiwanese accent, /iɔ̃/ is marginal/unnecessary, at most occurring only in onomatopoeia. However, speakers of Zhangzhou-style accents such as in Yilan systematically use either /iɔ̃ ~ iõ/ instead of the rime /iũ/.

Nasal Coda Rimes

For some accents, there is some variation between which words use /iɔŋ/ and which use /iaŋ/, with Zhangzhou-style accents having more /iaŋ/.

In Peh-oe-ji and Tai-lo, note that the <o> before codas -m and -ng indicates the open /ɔ/ vowel, rather than /o/ or /ə/.

Peh-oe-ji spells the /iŋ/ rime as <eng>, though this rime triggers palatalization of sibilant initials as with other rimes starting with /i/.

Checked Rimes

For Peh-oe-ji and Tai-lo, as with before the -m and -ng, the <o> before the -p and -ng indicates the open /ɔ/ vowel, rather than /o/ or /ə/. However, there is technically a distinction between /ɔ/ and /o ~ ə/ before the glottal stop coda /ʔ/, though the /ɔʔ/ sequence is limited to a small number of onomatopoeia/sentence-final particles or other colloquialisms.

Peh-oe-ji spells the /ik/ rime as <ek>, though this rime triggers palatalization of sibilant initials as with other rimes starting with /i/.

Nasalized Checked Rimes

As with in Chaoshanese, this category of nasalized checked rimes is characteristic of the vernacular register, and liked all checked rimes, they can only be pronounced in the Ru tones (Yinru and Yangru, tones 4 and 8). As with normal nasalized rimes, when coming after nasal initials m-, n-, ng-, the nasalized part of the rime is omitted from the spelling, so minnh > mih.

Tones

Citation Tones

Taiwanese contrasts five tones in unchecked syllables and two tones for checked syllables, for a total of seven tones. The Ru tones (Yinru and Yangru) are for checked syllables (syllables with ending -p, -t, -k, -h), the other six tones (Ping, Shang, Qu tones) are for non-checked syllables.

The neutral tone can occur as an indistinct, neutralized low tone in non-initial position.

Tone Sandhi

Whenever in non-final position of an utterance, the tone of a syllable is subject to changing into another tone contour. In Taiwanese, the changes are as follows:

Yinping (Tone 1) changes to Yangqu (Tone 7).

Yinshang (Tone 2) changes to Yinping (Tone 1).

Yinqu (Tone 3) changes to Yinshang (Tone 2).

Yangping (Tone 5) changes to either Yangqu (Tone 7) or to Yinqu (Tone 3). Change to Tone 7 is more typical of the south, and change to Tone 3 is more typical of the north.

Yangqu (Tone 7) changes to Yinqu (Tone 3).

Yinru (Tone 4) and Yangru (Tone 8) swap tone values with each other (Yinru turns into Yangru and Yangru turns into Yinru) if the checked coda is -p, -t, or -k.

Yinru with coda -h changes to Yinshang (Tone 2), and Yangru with coda -h changes to Yinqu (Tone 3).

If a syllable is followed by all neutral tone syllables, then it is counted as the final syllable of its utterance, and therefore will be in the original citation tone.

Consonants in Initial Position

Consonants in initial position can be grouped together into several classes:
plain:
tense:
aspirated:
sonorant:

Consonants Changes in Intervocalic Position

Final Position Consonants

Consonants in Final Position