At a Glance

Mandarin Chinese has more native speakers than any other language on Earth. Over 900 million people speak it natively, and tens of millions more use it as a second language across mainland China, Taiwan, Singapore, Malaysia, and a worldwide diaspora.

The standard form has three names. In mainland China it is Pǔtōnghuà (普通话, "common speech"). In Taiwan, Guóyǔ (國語, "national language"). In Singapore and Malaysia, Huáyǔ (华语, "Chinese language"). These are the same standard. Minor differences exist in pronunciation, vocabulary, and which characters get written, but a speaker of one understands the others without trouble.

Calling this language simply "Chinese" is convenient but technically misleading. Chinese is not a single language. It is a branch of the Sino-Tibetan family with several major members: Mandarin, Cantonese (Yue), Hokkien (Min Nan), Hakka, Wu (Shanghainese), Gan, and Xiang. They share a writing system and a long literary tradition, but a Mandarin speaker and a Cantonese speaker generally cannot understand each other in conversation. When people say "Chinese" they usually mean Mandarin, because it is the official spoken standard of the People's Republic of China, the Republic of China (Taiwan), and Singapore.

Two features stand out when you first encounter the language. The first is tone. Every full syllable carries one of four pitch contours plus a neutral tone. Change the pitch and you change the word: mā 妈 is "mother", má 麻 is "hemp", mǎ 马 is "horse", mà 骂 is "scold". The second is the near absence of morphology. Verbs do not conjugate. Nouns do not change for number or case. Mandarin has no past tense. Grammatical relations are encoded by word order, particles, classifiers, and context.

Varieties

Within Mandarin itself, dialectologists distinguish four sub-zones. Northern Mandarin (北方话) covers Beijing and most of the North China Plain. It is the basis of the standard, with retroflex consonants (zh, ch, sh) and a prominent feature called erhua, where the suffix -儿 gets attached to nouns. Northwestern Mandarin (西北官话) is spoken across Gansu and Shaanxi. Southwestern Mandarin (西南官话) covers Chengdu, Kunming, and Wuhan. That is tens of millions of speakers, and many of its varieties are not mutually intelligible with Beijing speech. Lower Yangtze (Jianghuai) Mandarin centers on Nanjing and Hefei.

Standard Mandarin is not the same thing as the Beijing dialect. The standard was codified in stages. It was first established as Guóyǔ in 1932, then renamed Pǔtōnghuà in the 1950s. The codifiers took Beijing pronunciation as their phonological model, Northern Mandarin as their dialectal basis, and modern vernacular writing as their grammatical norm. Real Beijing speech has features the standard does not: heavier erhua, additional tone sandhi rules, and a recognizable colloquial register where alveolar sibilants are articulated nearer the teeth and final -n and -ng are weakened or lost. Educated speakers shift between the standard and these local features depending on context.

The biggest regional split today is between mainland Pǔtōnghuà and Taiwan's Guóyǔ. Pronunciation is one difference. Taiwan tends to merge the retroflex zh-, ch-, sh- with z-, c-, s-, and erhua is rare. Vocabulary is another. 网络 wǎngluò versus 網路 wǎnglù for "internet". 软件 ruǎnjiàn versus 軟體 ruǎntǐ for "software". 出租车 chūzūchē versus 計程車 jìchéngchē for "taxi". Writing is a third. Mainland China uses Simplified characters, Taiwan uses Traditional. Singaporean and Malaysian Huáyǔ has its own character. Older speakers blend Hokkien, Cantonese, Teochew, and Bazaar Malay vocabulary into their Mandarin. Younger speakers, shaped by the post-1979 Speak Mandarin Campaign, code-switch heavily with English instead.

Mandarin's other axis of variation is register. There is no grammatical politeness system of the kind English has with its subjunctive. Instead, register is encoded lexically and pragmatically. Pronouns shift. 你 nǐ is the ordinary "you"; 您 nín is the honorific form. Formal and colloquial pairs cover everyday vocabulary. 诸位 zhūwèi versus 大家 dàjiā for "everyone". 别人 biérén versus 人家 rénjiā for "other people". 自己 zìjǐ versus 自个儿 zìgěr for "oneself". A layer of classical-register grammar still surfaces in formal writing. Public notices use 勿 wù, 莫 mò, or 禁止 jìnzhǐ for "do not", where ordinary speech uses 别 bié or 不要 bú yào. The formal passive marker 被 bèi gets replaced in speech by 让 ràng, 叫 jiào, or 给 gěi. Both registers are real Mandarin, picked up in different settings.

How it works

Mandarin's basic word order is subject–verb–object, but the language is **topic-prominent**. The most important information often comes first, even when that means the object precedes the verb. 这本书我看过 zhè běn shū wǒ kànguo means "This book, I have read." Inside the noun phrase, every modifier comes before the noun it describes. Adjectives, possessors, relative clauses, demonstratives. All of them precede the noun and are joined by the particle 的 de.

There is almost no inflection. Verbs do not change for tense, person, or number. Mandarin has no past tense. Time is supplied by adverbs like 昨天 zuótiān "yesterday" and by aspect particles attached to the verb. 了 -le marks an event as bounded. 过 -guo marks experiential aspect, meaning "have done at least once." 着 -zhe marks an ongoing state. Preverbal 在 zài marks an event in progress. Modality piles on through a set of auxiliaries. 会 huì, 能 néng, 可以 kěyǐ, 要 yào, and 应该 yīnggāi all stack before the verb. Sentence-final particles like 吗, 呢, 吧, 啊, 了, and 啦 close the clause, turning it into a question, a suggestion, or a notification of changed circumstances.

Two construction-level features stand out. The 把 (bǎ) construction lets a speaker move a definite object in front of the verb. It marks what linguists call a "disposal." The verb does something to that object. 我把书看完了 wǒ bǎ shū kàn-wán le: "I finished reading the book." The 被 (bèi) construction is the formal passive. 这本书被他买走了 zhè běn shū bèi tā mǎi-zǒu le: "The book was bought by him." And then there is the famous trio of homophones: 的, 地, and 得. All three are pronounced identically as de. They are written with different characters and play different grammatical roles. 的 links a modifier to a noun. 地 links a manner adverbial to a verb. 得 introduces a verbal complement.

Counting things requires a classifier. You cannot say "three book." You say 三本书 sān běn shū, literally "three book-classifier book." There are about fifty common classifiers and a few hundred specialized ones. 条 tiáo for long thin things. 张 zhāng for flat things. 只 zhī for animals. And the all-purpose 个 gè covers most everyday classifier use.

The writing system is logographic. Each character, 字 zì, represents a syllable that is also, in most cases, a meaningful morpheme. About 3,500 characters cover most general-purpose reading. Roughly 80% of the inventory consists of phono-semantic compounds. A meaning radical gets glued to a phonetic component. Take 妈 mā "mother." It combines 女 "woman" with 马 mǎ "horse," where the horse component is read for its sound. Mainland China and Singapore use Simplified characters, the product of reforms between 1956 and 1964 that streamlined around 2,200 character forms. Taiwan, Hong Kong, and Macau retain Traditional characters.

Mandarin also has an alphabet alongside its characters. Hanyu Pinyin was made the PRC's official romanization in 1958 and became the ISO international standard in 1982. It transcribes every Mandarin syllable in the Latin alphabet, using four diacritics for the four lexical tones: mā, má, mǎ, mà. The neutral tone is unmarked. Pinyin is how children on the mainland learn to read. It is how adults type on their phones. It is how this article spells Mandarin words. Taiwan uses a different system called Bopomofo (注音符号), a non-Latin phonetic script of 37 symbols and tone marks, often used alongside Pinyin.