← Chinese Pronunciation Guide

z c s vs zh ch sh: Flat-Tongue and Retroflex Sounds Told Apart

ChinesePath · Updated September 8, 2026

Two trios of Mandarin initials form a perfect mirror system: the flat-tongue (平舌) sounds z, c, s and the retroflex (卷舌) sounds zh, ch, sh. Within each trio the three members differ in how the air moves; between the trios, the difference is a single variable — where the tongue tip is. Get that one contrast into muscle memory and six initials fall into place at once, along with famous minimal pairs like 山 shān (mountain) vs 三 sān (three) and 猪 zhū (pig) vs 租 zū (to rent).

One difference: the tongue tip's address

• z, c, s — the tongue tip touches (or nearly touches) the upper teeth/gum ridge, right behind the front teeth. Flat tongue, front position.
• zh, ch, sh — the tongue tip curls up and back so that the front of the tongue meets the hard palate. That curl is what "retroflex" means, and it gives these sounds their darker, deeper color.

The manner of airflow then pairs up exactly across the two trios:

Two warnings on the anchors. First, pinyin c is /ts/ — never the English "k" or "s" values of the letter. Second, the English sh of she is not the Chinese sh — it is actually closer to Chinese x (see zh ch sh vs j q x); the Chinese sh requires the retroflex curl. The fourth retroflex, the voiced r, has its own article.

The special i: zhi, chi, shi, zi, ci, si

After these initials (plus r), the letter i is not "ee". It is a special final: the syllable just sustains the initial's own sound — tongue frozen in place, voice on. 知 zhī, 吃 chī, 师 shī, 资 zī, 次 cì, 思 sī all end in a buzzing or hissing prolongation, not in the vowel of 你 nǐ. Only after b, p, m, d, t, n, l, j, q, x is written i the true "ee" vowel.

Minimal pairs to train the contrast

Same final, same tone — only the tongue position distinguishes the words:

• 猪 zhū (pig) — 租 zū (to rent)
• 出 chū (to go out) — 粗 cū (thick, coarse)
• 山 shān (mountain) — 三 sān (three)
• 晒 shài (to sun-dry) — 赛 sài (competition)
• 主要 zhǔyào (main) — 组合 zǔhé (combination)
• 山体 shāntǐ (mountain body) — 三体 sāntǐ (Three-Body)

Drill method: say the flat-tongue member with the tongue tip planted at the teeth, then the retroflex member with the tip curled clearly back — exaggerate the gap at first, then shrink it to natural size. Careful listening and imitation of contrasted pairs beats tongue twisters, which mostly train frustration.

Do native speakers always distinguish them?

In the standard language, strictly yes — the contrast is fully phonemic. In many southern accents the retroflexes flatten toward z/c/s, and Chinese speakers still understand each other from context. For learners, the standard contrast is worth acquiring: it is what dictionaries, pinyin input and listening materials assume, and hearing it is as important as saying it — 时事 shíshì (current affairs), 事实 shìshí (fact), 历史 lìshǐ (history) and 四十 sìshí (forty) tangle quickly for ears that ignore the difference.

Frequently Asked Questions

What is the difference between zh ch sh and z c s?

Tongue position only. z/c/s put the tongue tip at the upper teeth (flat-tongue); zh/ch/sh curl the tip up to the hard palate (retroflex). The airflow types pair up one-to-one across the sets: zh/z unaspirated, ch/c aspirated, sh/s frictional.

How is pinyin c pronounced?

As /ts/ with a strong puff of air — like the ts of cats but at the start of a syllable: 菜 cài, 次 cì, 粗 cū. It is never the English "s" or "k" reading of the letter c.

Why don't zhi, chi, shi, zi, ci, si sound like "jee/chee/see"?

Because after zh/ch/sh/r/z/c/s the letter i is a special final: the tongue stays in the initial's position and the sound is simply voiced and prolonged. 知 zhī is a sustained zh-buzz, 思 sī a sustained hiss — no "ee" vowel.

Is Chinese sh the same as English sh?

No. English sh (she) is made further forward and is actually closer to Chinese x (西 xī). Chinese sh needs the tongue tip curled back to the hard palate — the retroflex posture — giving it a deeper, darker sound (师 shī).

How can I practice telling them apart?

Minimal pairs with everything equal except the initial: 猪 zhū / 租 zū, 出 chū / 粗 cū, 山 shān / 三 sān. Alternate slowly, exaggerating the tongue positions, and listen to contrasted recordings — targeted contrast work outperforms tongue twisters.