← Back
Reference Guide

Tibetan Spelling Rules

Tibetan syllables follow strict structural rules. This guide explains the components of a syllable and the rules the spellchecker enforces.

Anatomy of a Tibetan Syllable

Every Tibetan syllable is built around a single root letter. Up to six additional components may surround it, each occupying a defined position. Not all positions are filled — most syllables use only two or three.

PrefixSuperscriptRootSubscriptVowelSuffixPost-suffix
before rootabove rootroot letterbelow rootabove or below stackafter rootfinal
བསྒྲུབ(bsgrub — to accomplish)

Prefix (སྔོན་འཇུག)

Placed before the root. Only five letters can be prefixes:

Not every combination of prefix + root is valid. The spellchecker knows which combinations are attested in classical Tibetan.

Superscript (མགོ་ཅན)

Sits above the root. Only three letters can be superscripts:

Each superscript only combines with specific roots. For example, ས can sit above ཀ (སྐ) but not above ར.

Subscript (ཞབས་ཀྱུ)

Written below the root consonant. Four possible subscripts:

Like superscripts, each subscript only combines with certain roots. For example, ya-btags (ྱ) can appear under ཀ (ཀྱ) but not under ང.

Vowel

Marked above the stack. The default vowel is a (unmarked).

gigu
shabkyu
drangbu
naro

Only one vowel mark is allowed per syllable. Two consecutive vowel marks is an error, with a few notable exceptions.

Suffix (རྗེས་འཇུག)

Follows the root. Ten possible suffixes:

The suffix determines which particle forms are valid after the word (see Particle Rules below).

Post-suffix (ཡང་འཇུག)

The final element, appearing after the suffix. Only two are valid:

Root is required. Every valid Tibetan syllable must have a root consonant. All other components are optional.

Particle Rules

Particles are small grammatical words that follow nouns and verbs. In classical Tibetan, a particle's spelling changes depending on the last letter of the word it follows. This is called euphonic change — the forms are chosen to make pronunciation flow naturally.

The determining factor is the suffix of the preceding word (or the absence of a suffix).

Relational Particle (of, 's)

Marks possession or association. The correct form depends on the preceding word's suffix:

ParticleUsed after suffixExample
ཀྱིད བ སབོད་ཀྱི་of Tibet
གིག ངཆོས་གི་of the dharma
གྱིན མ ར ལསེམས་ཅན་གྱི་of beings
འིའ or no suffixལྷའི་of the deity
ཡིlenientany (lenient variant)ལྷ་ཡི་used in poetry for metrical reasons

Agentive Particle (by, with)

Marks the agent, instrument, reason, or source of an action — roughly equivalent to “by,” “with,” or “because of.” The forms follow the same suffix groupings as the relational particle:

ParticleUsed after suffixExample
ཀྱིསད བ སབོད་ཀྱིས་by Tibet
གིསག ངཆོས་གིས་by the dharma
གྱིསན མ ར ལསེམས་ཅན་གྱིས་by beings
འ or no suffixལྷས་by the deity
ཡིསlenientany (lenient variant)used in poetry for metrical reasons

Locative Particle (in, at, to)

Marks location or direction. The locative has more forms than the other particles:

ParticleUsed after suffixExample
any suffix or no suffixབོད་ན་in Tibet
any suffix or no suffixབོད་ལ་to Tibet
སུཆོས་སུ་into the dharma
ཏུག བ (or post-suffix ད)ཕྱོགས་ཏུ་in the direction
དུང ད ན མ ར ལབར་དུ་until
རུའ or no suffixརྒྱལ་ཁབ་རུ་in the kingdom
ཆེར་greatly

Indefinite Article (a, one)

The indefinite article / number one. Three forms based on the preceding suffix:

ParticleUsed after suffixExample
ཅིགག ད བཆོས་ཅིག་a dharma
ཤིགཕྱོགས་ཤིག་a direction
ཞིགང ན མ ར ལ འ or no suffixལྷ་ཞིག་a deity

Error Types

When the spellchecker finds a problem, it identifies the type of error. Here is what each type means:

Wrong particle form

བོད་གི་

The particle doesn't match the suffix of the word before it. Here གི་ is used after བོད་ (which ends in ད), but ད calls for ཀྱི་.

Fix: Use ཀྱི་ after words ending in ད, བ, or ས.

Invalid prefix combination

ཧཧིབ་

The letter before the root is not a valid prefix for that root. ཧ (ha) cannot act as a prefix.

Fix: Only the five prefix letters (ག ད བ མ འ) can precede a root, and only in attested combinations.

Invalid superscript combination

A superscript letter is stacked above a root it cannot combine with. Each of the three superscripts (ར ལ ས) only stacks above specific roots — for example, ས can sit above ཀ to form སྐ, but not above ར.

Fix: Check which roots are valid for that superscript.

Invalid subscript combination

ངྱི

The subscript below the root is not valid for that root. ང (nga) cannot take a ya-btags subscript.

Fix: Subscripts (ྱ ྲ ླ ྭ) only combine with certain roots.

Invalid suffix

ཀཝ་

The letter after the root is not one of the ten valid suffixes.

Fix: Only these ten letters can be suffixes: ག ང ད ན བ མ འ ར ལ ས.

Double vowel

Two vowel marks appear on the same syllable, which is structurally impossible.

Fix: Each syllable can only carry one vowel mark.

Encoding error (critical)

wrong character

A character that looks like Tibetan is actually the wrong Unicode codepoint — a common issue when text is copied from certain older fonts or systems.

Fix: Re-type the syllable from scratch using a Unicode Tibetan keyboard, or use a font conversion tool.