The Third Tone Is Usually Just Low

Textbooks introduce the 3rd tone as a dramatic valley — the voice dips low and then swoops back up, drawn as a wide U. That description is true for a syllable said completely on its own, and it is what you hear on a vocabulary app playing one word at a time. But it is not how the 3rd tone usually sounds in a real sentence, and copying that full swoop onto every 3rd tone is one of the surest signs of a foreign accent. In connected speech the 3rd tone is, most of the time, simply low — it sinks to the bottom of your range and stays there. Getting this right instantly makes your Mandarin sound smoother and more native. This builds directly on the four tones; make sure you can produce all four in isolation first.

The "half-third": low and flat

Phoneticians call the everyday version the half-third tone. The name captures exactly what happens: the syllable keeps the first half of the textbook contour — the drop to the bottom of your voice — and drops the second half — the rise back up. So instead of the full 2-1-4 valley, you get a short, low, roughly level tone that sits at pitch 1 or 2 and goes nowhere.

This half-third is what you use whenever a 3rd tone is followed by another syllable — that is, almost always, since most 3rd-tone syllables live inside words and sentences rather than alone.

好人

hǎorén

a good person — 好 stays low and flat; there is no rise before 人.

我们

wǒmen

we / us — 我 is a low half-third; do not lift it back up.

老师

lǎoshī

teacher — 老 sinks low and holds; 师 then rides on top with its 1st tone.

In all three, the 3rd-tone syllable does the dip and skips the rise. If you swoop up on hǎo in hǎorén, you insert a little bump that native speakers don't make — and worse, the rise can make the listener hear a different tone.

💡
Think of the 3rd tone's real job as going low, not going down-and-up. The low target is the identity of the tone; the final rise is a bonus that only appears when the syllable has room to breathe at the end of a phrase.

When the full dip does come back

The complete dipping-and-rising contour is not a myth — it is just reserved for two specific situations:

  1. The syllable is said alone, e.g. as a one-word answer or an exclamation.
  2. The syllable ends a phrase or sentence, right before a pause, where there is nothing following to cut the contour short.

好!

hǎo

OK! / Great! (informal) — said alone, so the full low-then-rise contour comes out.

你几岁?——五。

Nǐ jǐ suì? — Wǔ.

How old are you? — Five. (informal) — 五 is a one-word answer, so it gets the full dip.

我姓李。

Wǒ xìng Lǐ.

My surname is Li. (neutral) — 李 ends the sentence, so it can take the fuller contour before the pause.

Compare that phrase-final with a mid-sentence 3rd tone, which stays a flat half-third. The same syllable behaves differently depending only on where it sits.

The half-third in real sentences

Here the half-third does most of the work, because 3rd tones keep landing before other syllables:

我有时间。

Wǒ yǒu shíjiān.

I have time. (neutral) — 我 and 有 are both low half-thirds running into what follows.

你是老师吗?

Nǐ shì lǎoshī ma?

Are you a teacher? (informal) — 你 and 老 stay low and flat, not swooping.

请坐。

Qǐng zuò.

Please, have a seat. (polite) — 请 is a low half-third leading straight into the falling 坐.

我很累。

Wǒ hěn lèi.

I'm tired. (informal) — 我 and 很 sit low; the energy lands on the falling 累.

Notice how in hěn lèi two 3rd tones in a row both stay low and understated, and the sentence's emphasis falls on the 4th-tone lèi. That low, un-fussy delivery of the 3rd tones is exactly what makes it sound relaxed and native rather than laboured.

💡
A useful mental image: a mid-sentence 3rd tone is a stepping stone you rest your foot on at the bottom, not a trampoline you bounce back up from. Touch down low, then move on to the next syllable.

The one exception you should know about now

There is a separate, very common rule for what happens when two 3rd tones bump into each other — as in 你好 (nǐ hǎo). There, the first 3rd tone doesn't become a half-third; it turns into a full rising 2nd tone (ní hǎo). That is third-tone sandhi, and it is important enough to learn as its own rule. For now, just know that "3rd tone before another 3rd tone" is the special case; "3rd tone before a 1st, 2nd, 4th, or neutral tone" is the low half-third described on this page.

How to compare source-language intuition

English has nothing quite like this, because English pitch belongs to whole phrases (intonation), not to individual syllables. But you can borrow an English feeling: the half-third is like the low, level pitch you land on inside a flat, slightly bored statement — "I guess so" — where "guess" sits low and doesn't rise. That low resting pitch, held steady, is your half-third. What English never does is add a little upward flick to every low syllable, which is exactly the habit you must avoid. If anything, English speakers should aim to make their 3rd tones lazier and lower than feels natural.

Common Mistakes

1. Performing a big U-shaped valley on every 3rd tone. This is the classic beginner giveaway. Said in a sentence, an over-swooped 3rd tone sounds slow, sing-song, and unnatural — and it eats time the sentence's rhythm doesn't have.

❌ 我们 pronounced wǒ↘↗men (full dip-and-rise on 我)

Incorrect — swooping back up on 我 sounds laboured and foreign.

✅ 我们 pronounced wǒmen (我 low and flat)

Correct — 我 just sinks low and runs into men.

2. Adding a rise that turns the 3rd tone into a 2nd tone. If you swoop up on a non-final 3rd tone, listeners may hear the rising 2nd tone instead — a real change of word. The half-third protects you by staying low.

❌ 老师 pronounced láoshī

Incorrect — rising on 老 makes it sound like the 2nd tone (a different, non-existent word here).

✅ 老师 pronounced lǎoshī with 老 low and flat

Correct — 'teacher', with the 3rd tone kept low.

3. Dropping the low target and keeping only a rise. The reverse error: some learners skip the low part entirely and just glide upward. But the low is the non-negotiable core of the 3rd tone — lose it and the tone is gone.

4. Forgetting to let the final 3rd tone breathe. At the end of a sentence or as a one-word answer, the fuller contour is natural and even helpful for clarity. Clipping it flat there ( with no dip at all) can sound abrupt. Low-and-flat is for the middle; the fuller shape is fine at the edge.

Key Takeaways

  • In connected speech the 3rd tone is usually a low, flat half-third — the dip without the rise.
  • Use the half-third before a 1st, 2nd, 4th, or neutral tone (i.e. almost everywhere).
  • The full dip-and-rise only appears when the syllable is said alone or ends a phrase.
  • The tone's identity is going low, not going down-and-up — don't over-perform the valley.
  • Two 3rd tones in a row are a different rule (third-tone sandhi): the first becomes a rising 2nd tone.

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 Four TonesHSK 1Mandarin has four full tones — high-level (mā), rising (má), low-dipping (mǎ), and sharp-falling (mà) — each a pitch shape you can map onto a familiar English pitch gesture; the tone is part of the word, not decoration.
  • Third-Tone Sandhi (你好 → ní hǎo)HSK 1When two 3rd tones collide, the first turns into a rising 2nd tone — 你好 is written nǐ hǎo but pronounced ní hǎo, and 很好 comes out hén hǎo; the pinyin keeps the original marks, so your mouth must apply the change automatically. This is the single most important sandhi rule to automate early.
  • The Neutral Tone (轻声)HSK 1Some syllables are toneless — the 轻声 (qīngshēng) — spoken short, light, and at a pitch handed to them by the syllable before; written with no tone mark (māma, xièxie, de, le, zi), it works like an unstressed English syllable and also marks particles and the noun suffix 子.
  • How to Practice TonesHSK 1Tones are a physical, trainable habit learned like a musical instrument — through many short focused reps with immediate feedback, not by reading rules; exaggerate the pitch shapes, trace them with your hand, anchor each tone to a mental gesture, record and compare against native audio, and always learn a new word together with its tones.