Pronunciation guide

How to read and say Cantonese

Letters, sounds and stress from Wikivoyage's Cantonese phrasebook. Then hear the phrases and practice them, free.

▶ Practice Cantonese phrases

Like other Chinese languages, Cantonese is written using Chinese characters but employs its own "unique" pronunciation for those characters.

The pronunciations given in this guidebook use the Yale Romanization system. Unlike in Mandarin, there is no widely-used Romanization system for Cantonese, and native speakers almost never learn those that do exist. Since most locals will have no idea how to read Romanized Cantonese, stick to Chinese characters for written communication.

Unless otherwise indicated, pronunciation between the Cantonese of Hong Kong and Macau is identical to that of Guangzhou.

Consonants

Like Mandarin, Cantonese distinguish between aspirated and unaspirated consonants, not unvoiced and voiced as in English, and lacks voiced consonants. Aspirated sounds are pronounced with a distinctive puff of air as they are pronounced in English when at the beginning of a word, while unaspirated sounds are pronounced without the puff, as in English when found in clusters. However, Cantonese lacks the "tongue rolling" (pinyin zh, ch, sh, r) initial consonants that are found in Mandarin.

Yale Pronunciation
b p in "sport"
p p as in "pat"
m m as in "mom"
f f as in "foot"
d t in "stop"
t t as in "top"
n n as in "not"; in many areas including Hong Kong and Macau, initial n is often substituted with l as in "lap"
l l as in "lap"
g k in "sky"
k k as in "kite"
ng ng as in "singer"; in Hong Kong and Macau, ng by itself without a vowel is often substituted with m as in "mom", while it is often omitted as the initial consonant of a syllable
h h as in "hot"
j blend of the z in "Mozart" and the j in "judge"
ch blend of the ts in "cats" and the ch in "church"
s s as in "sleep"
gw qu as in "square"
kw qu as in "quark"
y y as in "yard"
w w as in "want"; otherwise, like English "v" in "victory"

Finals

Unlike Mandarin, Cantonese retains all the final consonants (m, n, ŋ, p, t, and k) of Middle Chinese. The final consonants p, t, and k are unreleased. This means that they are virtually silent and you hear no "puff of air" at the end of the syllable.

Yale Pronunciation
aa a as in "spa"
aai igh as in "sigh"
aau ow as in "how"
aam am as in "Vietnam"
aan an as in "Taiwan"
aang combination of aa and ng
aap op as in "top" in General American, arp as in "tarp" in Received Pronunciation
aat ot as in "hot"in General American, art as in "tart" in Received Pronunciation
aak ock as in "rock"in General American, ark as in "bark" in Received Pronunciation
ai (IPA: ɐɪ̯) i as in "kite"
au ou as in "scout"
am ome as in "some"; otherwise, like "am" in "ham".
an un as in "sun". This can be pronounced like "an" in "man"
ang ung as in "lung". This may sound like "ang" in "rang".
ap up as in "cup". This can be pronounced like "ap" in "map".
at ut as in "cut"; otherwise, like "at" in "cat".
ak uc as in "suck"; otherwise, like "ack" in "back".
e (IPA: ɛ) e as in "bet"
ei ay as in "say"
em em as in "temple"
eng eng as in "penguin"
ek eck as in "peck"
i ee as in "tee"
iu ew as in "few"
im eem as in "seem"
in een as in "seen"
ing ing as in "sing"
ip eep as in "sleep"
it eet as in "meet"
ik ick as in "sick"
o aw as in "paw" (Received Pronunciation)
oi oy as in "boy"
ou oe as in "toe"
on orn as in "scorn" (Received Pronunciation)
ong ong as in "song" (Received Pronunciation)
ot ot as in "hot" (Received Pronunciation)
ok ock as in "stock" (Received Pronunciation)
u oo as in "coo"
ui ooey as in "gooey"
un oon as in "soon"
ung combination of ou and ng
ut oot as in "boot"
uk ook as in "book"
eu (IPA: ɵ) er as in "her" (Received Pronunciation, with rounded lips)
eung combination of eu and ng
euk ork as in "work" (Received Pronunciation)
eui o as in "no" (Received Pronunciation)
eun on as in "person"
eut ot as in "carrot"
yu (IPA: y) u as in "tu" (French)
yun un as in "une" (French)
yut No equivalent in English. Similar to a German jü followed by a t
m mm as in "hmm"
ng (IPA: ŋ̩) ng as in "sing"

Tones

Cantonese is a tonal language. This means that the same syllable, pronounced in a different tone, has a different meaning. To complicate this, there may be more than one character pronounced as the same syllable with the same tone. In this case, context usually helps resolve the ambiguity. This may sound daunting, but is in fact is better than say, English, where there are a great deal of words that are spoken identically (eg. their, there, they're) and have nothing but context to help determine which one it is. Cantonese has context and tone to help distinguish words.

Different variations of the Cantonese language have a different number of tones, from as few as six to as many as ten or more. The larger numbers involve distinctions that modern linguists don't consider tonal, so to get by, you only need to distinguish between the following six tones:

Yale Description Start-to-end pitch Yale Description Start-to-end pitch
1 ā High Level 4 àh Low Falling
2 á Mid Rising 5 áh Low Rising
3 a Mid Level 6 ah Low Level

The tonal pronunciation of Cantonese is by far the most difficult aspect of the often daunting language. The very minor initial difficulty in learning the tones is sometimes more than made up for by simple grammar, and absence of almost all plurals, genders, tenses and other conjugations that make many other world languages seem difficult by comparison.