INITIAL
zh: unaspirated retroflex affricate
Learn Mandarin zh: the tongue tip curls up behind the ridge, with no puff of air. See how it differs from j and z.
- Broad IPA
- [ʈʂ]
Mouth & tongue
Enable JavaScript to view the 3D panda. Pronunciation steps remain available below.
Vocal folds: vibrating
TongueTeethUvula
Simplified human pronunciation. The small uvula hangs from the palate; it is not the vocal folds. The wave indicates voicing.
Nasal: the soft palate lowers; air exits the nose
High and level, around 55 on the five-level pitch scale.
Voice settings
Voice
Finding available voices…
Optional eSpeak NG: a mechanical voice generated on your device. It downloads from this site only when you choose it, and reads pinyin from our dictionary. Its pitch range is narrow, so it is a fallback rather than a model for tones.
eSpeak NG · GPL-3.0 · SourceThe mouth starts slightly open. Movement is slowed for observation; the synthetic sound plays once at its original speed. Pause holds the current mouth shape; reset returns to the relaxed position. The panda is a symbolic teaching aid for human pronunciation, not animal anatomy. A simplified teaching model, not a measured or anatomically validated simulation. It shows the soft palate lowering for nasal sounds, but not airflow. Single syllables use eSpeak NG, a synthetic voice on your device. Its pitch range is narrower than natural speech, so use the tone diagrams and word buttons as well. Words and sentences use a voice from your browser or device when one is available. No native recordings are bundled.
Words to practice
Sentence to practice
Listen and compare
Tip back for the first word; tip forward behind the teeth for the second.
Pinyin here shows dictionary tones. Spoken tones change in some contexts, such as nǐ hǎo; see the lessons on tone changes.
What you are learning
Pinyin zh, ch, sh and r share a raised, drawn-back tongue tip. In much of southern China and Taiwan these are often said less retroflex; that is regional variation, not an error.
What your mouth is doing
The tip rises behind the alveolar ridge, stops the air, then releases with friction. There is little or no puff, and the lips stay relaxed unless the vowel rounds them.
Compared with English
English j in judge is the nearest cue, but it is further forward, often rounded and voiced. Pull the tip back and remove the buzz. Do not curl so far that the underside of the tongue is pressed against the palate.
Build the sound, step by step
- Say r in English red and notice the tip off the palate.
- Touch the tip behind the ridge to stop the air, and release softly: zhī.
- Try zhōngguó and zhù, with no puff.
What to watch for
- Moving the tip forward turns zhī into zī; that merger is common in some regions, but practice the difference.
- Adding air changes zhǐ into chǐ.
References & editorial notes
PronounceKit wrote this lesson as practical guidance for English speakers, based on the references below. The comparisons are editorial teaching suggestions. It describes common Mainland Standard Mandarin patterns, not every accent or regional variety, and it has not been independently reviewed by a pronunciation specialist or Mandarin teacher.
- Lee & Zee (2003), Standard Chinese (Beijing), Journal of the IPA
- Scheme for the Chinese Phonetic Alphabet (Hanyu Pinyin), 1958 — text on Wikisource
Published 5 November 2026