An initial is the consonant a syllable starts with — the m in 妈 (mā, "mom"), the h in 好 (hǎo, "good"). Mandarin has a tidy inventory of about 21 initials, and unlike the finals (the vowel part), most of them will feel familiar to an English speaker. The catch is a single, deep difference in how the sound system is organized: where English divides consonants into voiced and voiceless, Mandarin divides them into aspirated and unaspirated — with a puff of air or without. Retrain your ear for that one contrast and half of Mandarin's initials fall into place immediately.
The six families
Initials are easiest to learn in groups defined by where in the mouth they are made. Here is the whole set, organized by place of articulation.
| Family | Initials | Made where | English-ish anchor |
|---|---|---|---|
| Labials | b p m f | lips | spy · pie · my · fie |
| Alveolars | d t n l | tongue tip on the gum ridge | sty · tie · nigh · lie |
| Velars | g k h | back of the tongue | sky · kite · (loch) |
| Palatals | j q x | tongue blade at the hard palate | jeep · cheese · (sheee) |
| Retroflexes | zh ch sh r | tongue tip curled back | jump · church · shoe · leisure (curled) |
| Sibilants | z c s | tongue tip at the upper teeth | kids · cats · sea |
The labials, alveolars, and velars are the easy tier — they map closely onto English sounds. The palatals, retroflexes, and sibilants are the challenging tier, each with its own dedicated page: the retroflexes zh/ch/sh/r, the flat sibilants z/c/s, and the palatals j/q/x. This page gives you the framework and the one insight that unlocks the whole grid.
妈妈爸爸都很忙。
Māma bàba dōu hěn máng.
Mom and Dad are both busy. — labials m and b in one breath. (informal)
弟弟很喜欢喝牛奶。
Dìdi hěn xǐhuan hē niúnǎi.
My little brother loves drinking milk. — alveolar d, n, l plus velar h. (informal)
The one insight: aspiration, not voicing
In English, b and p are different because your vocal cords buzz for b (voiced) and stay silent for p (voiceless). Put a hand on your throat: buzz for "buh," silence for "puh." That voicing contrast is the backbone of the English consonant system.
Mandarin does not work this way. Neither Mandarin b nor Mandarin p is truly voiced — both are voiceless. What separates them is aspiration: the little puff of breath that follows the release.
- b = an unaspirated [p] — lips release with no puff of air.
- p = an aspirated [pʰ] — lips release with a strong puff of air.
The same split runs through every stop pair: d/t, g/k, and (on their own pages) j/q, zh/ch, z/c. In each pair the first is unaspirated, the second aspirated.
Here is the good news for English speakers: you already make unaspirated stops — you just don't notice. Compare the p in "pin" (a big puff) with the p in "spin" (no puff). The p after s is exactly Mandarin b. So to say 爸 (bà), whisper "spa" and drop the s: the leftover pa is a perfect Mandarin bà. Do the same with "star" → d, "sky" → g.
爸爸怕水。
Bàba pà shuǐ.
Dad is afraid of water. — 爸 (bà, no puff) vs 怕 (pà, big puff) side by side: the aspiration is the whole contrast. (informal)
Hearing the pairs: unaspirated vs aspirated
These three minimal pairs differ only in aspiration — same vowel, same tone. Train your ear on them until the puff of air, not any imagined voicing, is what you hear.
白
bái
white — b, unaspirated (like the p in 'spy').
排
pái
row, line — p, aspirated (like the p in 'pie'). Same as 白 but with a puff.
都
dōu
all, both — d, unaspirated (the t in 'sty').
偷
tōu
to steal — t, aspirated. 都 dōu vs 偷 tōu differ only by the puff.
该
gāi
should, ought to — g, unaspirated (the k in 'sky').
开
kāi
to open — k, aspirated. 该 gāi vs 开 kāi: same vowel, same tone, only aspiration differs.
The three that need special notes
Most initials behave, but three deserve a flag even at this overview level.
f is straightforward — the English f. m, n, l are the English sounds. But h is not quite the English h: standard Mandarin h is a velar fricative, made at the back of the tongue with audible friction, closer to the ch in Scottish "loch" or German "Bach" than to the soft breathy English "h." In fast speech it softens, but aim a little rougher and further back than you would in English.
我很喜欢喝红茶。
Wǒ hěn xǐhuan hē hóngchá.
I really like drinking black tea. — four h's; keep them slightly rough and back-of-the-mouth. (informal)
r at the start of a syllable (日 rì, 人 rén) is a retroflex sound, nothing like an English "r" — it lives with the retroflexes. And the semivowels y and w (as in 要 yào, 我 wǒ) are not counted among the 21 true initials; they are spelling devices for syllables that begin with an i-, u-, or ü-glide, covered with the finals.
Common Mistakes
1. Voicing b, d, g like English. Learners let their vocal cords buzz on 爸 (bà), 大 (dà), 高 (gāo), producing a heavy English "b/d/g." It's not fatal — natives key on aspiration, so you'll be understood — but it sounds foreign. Aim for a silent, puff-less release.
大
dà
big — a voiceless, unaspirated [t] (the t in 'sty'), NOT a buzzing English 'd'.
2. Aspirating b, d, g so they merge with p, t, k. This is the dangerous error, because it collapses meaning. If 爸 (bà) picks up a puff, it becomes 怕 (pà) — "Dad" turns into "fear."
爸
bà
dad. ❌ 'pà' (puff added) = 怕 'fear' → ✅ 'bà' (no puff).
3. Making h a soft English 'h'. Mandarin h has real friction at the back of the mouth. A too-soft h makes 喝 (hē, "drink") and 很 (hěn, "very") sound weak and unclear.
喝
hē
to drink — a velar fricative [x], rougher and further back than English 'h'.
4. Trying to "voice" the aspirated ones for contrast. Some learners overcorrect and make p, t, k voiced-and-puffed. They should stay voiceless — just add the air.
开
kāi
to open — voiceless k with a strong puff of air; don't buzz the throat.
5. Treating y and w as consonant initials to be memorized. They are glide spellings, not part of the 21-initial grid — learn them with the finals instead.
我要。
Wǒ yào.
I want (it). — 我 and 要 begin with glides w-/y-, not true initials. (informal)
Key Takeaways
- Mandarin has ~21 initials in six place-based families: labials b p m f, alveolars d t n l, velars g k h, palatals j q x, retroflexes zh ch sh r, sibilants z c s.
- The stop pairs contrast by aspiration, not voicing: b d g are unaspirated (no puff), p t k are aspirated (puff). English already has the unaspirated sounds after s (spy, sty, sky).
- The safe error is under-voicing; the meaning-breaking error is adding a puff to b/d/g and merging them with p/t/k.
- h is a back, slightly rough velar fricative, not the soft English "h."
- The hard families — retroflexes, sibilants, palatals — each get a dedicated page.
Now practice Chinese
Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.
Start learning Chinese→Related Topics
- The Retroflex Initials (zh ch sh r)HSK 1 — zh, ch, sh, and r are made with the tongue tip curled up and back toward the ridge behind the teeth — the sounds that most define a Mandarin accent — with zh/ch an unaspirated/aspirated affricate pair, sh a curled 'sh', and r a buzzy curled sound found in no English word.
- The Flat Sibilants (z c s)HSK 2 — z, c, and s are flat dental sounds made with the tongue tip near the upper teeth — z is 'dz' as in 'kids', c is 'ts' as in 'cats', and s is a plain 's' — with z/c contrasting by aspiration, and the letter c being the great beginner trap since it is never a 'k' or an 's'.
- The Palatals (j q x)HSK 1 — j, q, and x are made with the tongue blade spread against the hard palate and always take an i- or ü-type vowel — j like 'j' in 'jeep', q like 'ch' in 'cheese', x a hissy 'sh' with the tongue forward — and because they only precede i/ü, the ü drops its umlaut (ju = jü, qu = qü, xu = xü).
- Finals OverviewHSK 1 — A final is everything in a syllable after the initial — a main vowel, optionally a glide before it and a nasal (-n / -ng) or -r after it — and Mandarin has around 35 of them, built from just six vowel letters (a o e i u ü); the crucial shift for English speakers is that finals are learned as whole gliding units, not spelled out letter by letter, so the final in 天 (tiān) sounds like 'yen', not 'tee-AHN'.