Sổ tay Phát âm — the conversion chart

Learn the differences in these four books and you can pronounce any written Vietnamese word, including ones you have never seen and do not understand.

That claim would be a lie about English. Nobody can read "ghoti", "colonel" or "Worcestershire" from first principles; English spelling has to be learned word by word for a decade. Vietnamese spelling was designed, comparatively recently, to be read. One spelling, one sound. No silent letters. No exceptions list. Once you have the table, the table is the whole language.

So this is not a course. It is a conversion chart — from the reading you already do fluently, to the reading Vietnamese asks for. Your starting point is not zero. It is a literate adult applying the wrong sound values to a familiar alphabet.

The honest limit of the promise: reading is solved, spelling is not. In Hà Nội, d, gi and r all sound the same, as do s and x, and tr and ch. So you can always go from the page to the sound, but hearing a word will not always tell you how to write it. That is a real gap and it is one native Northerners have too.

Read this table and you have the 80%. The four books are the other 20%, plus the reasons.

You do not read these in a classroom order. Each one fixes a different part of the syllable, and the map below is the whole plan.

Books 1 and 2 are the ones that change whether you are understood at all. Book 3 is the one that changes whether you sound like a reader or a speaker. Book 4 is the one that never stops paying.

• Say everything out loud. Every book ends with a "prove it to yourself" table. Cover the answers, say the words, then check. Reading the answers is re-reading; producing them first is retrieval, and only one of those two builds anything.
• In the app, hold any Vietnamese word to hear it (highlight it on a laptop). Do that before you decide you have it right. You cannot reliably hear your own errors from the inside.
• Do not try to finish. Each book is a reference you re-enter, not a chapter you complete. The tables exist so that in three months you can look up one rhyme in ten seconds.

Everything in these books is Northern (Hà Nội) unless a section says otherwise. That is the app's default production dialect and the accent taught in Vietnamese schools and used in national media.

Each book ends with a short "you may also hear" section for Southern — kept separate on purpose, because mixing dialects is how learners end up with an accent from nowhere. Learn to produce one; learn to understand the others.

A one-line summary of the whole dialect map, which is worth carrying around:

The North merges consonants and keeps all six tones. The South merges tones and endings, and keeps its consonants apart. Between them nothing is lost, which is why the spelling has to record all of it.

Central (Huế, Đà Nẵng) is treated by FluentViet as a comprehension skill, not a target accent — see [[../cheatsheets/central-dialect]].

The claim at the top is checked, not asserted. coverage.py in this folder pulls every Vietnamese syllable out of the app's own scene and story content, decomposes each one into onset, rhyme and tone, and asserts that all three are covered by inventory.json, which in turn must appear in the prose of these four books.

Run it:

python3 01-language/pronunciation-guide/coverage.py

At the time of writing it reports 100% of onsets, rhymes and tones in the app's content covered, plus the rest of the standard Vietnamese inventory listed for completeness. The syllables it cannot parse are all English loanwords, brand names, digits and texting abbreviations — real Vietnamese text, but not Vietnamese orthography, and no pronunciation guide can or should cover wifi.

Back to the start of your journey — open the app's Guidebook and pick the book you need.