Literary and colloquial readings in Shanghainese

Sooner or later you meet a character you already know, pronounced a way you have never heard. You have not misremembered it. A good many Chinese characters carry two pronunciations in the same variety, and Shanghainese has more of them than Mandarin does.

Updated 20 September 2026 · 5 min read

Tap 🔊 to hear an example. Shanghainese examples play the recordings from Verbavia’s course, made with a Shanghainese voice; Mandarin examples play a Mandarin voice.

Two readings, two histories

The phenomenon has a name in Chinese, 文白异读 — literally literary and vernacular divergent reading. A character has a 白读, a colloquial reading, and a 文读, a literary one. They are not interchangeable and they are not free variation; each belongs to particular words.

The reason there are two is that they came from different places at different times. Wikipedia’s summary of the mechanism is worth having exactly: colloquial readings “are usually older, resembling the sound systems described by old rime dictionaries like the Guangyun, whereas literary readings are often closer to the phonology of newer sound systems”. The colloquial layer is what the language inherited and kept. The literary layer is what it borrowed, later, from whichever prestige variety mattered at the time — for northern Wu, the Beijing and Nanjing speech of the Ming and Qing, and then modern standard Mandarin.

So a doublet is a fossil of a borrowing. The everyday reading is the native one and the bookish reading is the import, which is the opposite of what the names suggest to an English speaker.

The clearest example, and it is in the course

Take 物 veh ‘thing, matter’. Now take the two ordinary words it appears in:

• 物事 meh zy ‘stuff, belongings’ — the everyday word, and the character is read meh.
• 事物 zy veh ‘things, matters’ — the bookish word, and the same character is read veh.

Same character, two readings, and the two words are near-anagrams of each other. 物事 meh zy ‘stuff’ is what is in your bag; 事物 zy veh ‘matters’ is what an essay is about. The colloquial meh survives in the homely word and the literary veh turns up in the abstract one, which is the rule in miniature.

This pair is worth memorising because it is the one place the distinction is visible with no special knowledge: both words are common, both are in the thousand-word list, and you can hear the difference.

The others you will meet

Below, the reading this course teaches is given first, because that is the one you will hear in its recordings. The second reading is the other member of the pair. Where a reading is not the course’s, it is set in italics rather than dressed up as something you can play.

A caution about that table. Which reading is “the” one for a given character depends on the word it is in, on the speaker’s age, and on the dictionary you consult, and sources disagree in the details. What is not in dispute is that the pairs exist and that they pattern the way the 物事 / 事物 example shows.

The rule of thumb

There is no formula, but there is a strong tendency, and it is worth carrying:

• Colloquial readings live in old, everyday, concrete words — the body, the house, food, family, the things people talked about before anyone wrote them down.
• Literary readings live in compounds, technical and abstract vocabulary, institutional names, and anything that arrived through writing or through Mandarin.

Which means the newer and more official a word feels, the likelier it takes the literary reading — and that a word borrowed from Mandarin last year will almost certainly take it, because the literary layer is where Mandarin has always entered this language.

Wu has many, Mandarin has few

This is not evenly distributed across Chinese. The Chinese-language account of the phenomenon ranks it plainly: the Min varieties have the most complex literary and colloquial systems, Wu and Xiang are intermediate, and Beijing Mandarin and Cantonese — the two dominant prestige varieties today — have comparatively few.

There is a logic to that. A variety that has spent centuries being the prestige standard has little to borrow prestige readings from; a variety that has spent centuries next to one accumulates layers. Shanghainese is in the second position, which is the same reason it differs from Mandarin in so many other ways — see the comparison page.

It is being flattened

Qian Nairong, the dialectologist most associated with documenting Shanghainese, has tracked the distinction eroding by generation. In his data, words such as 文学 (literature) and 男生 (male student) took the literary reading among speakers of the 1940s-to-1960s cohort and the colloquial reading among those of the 1970s-to-1990s cohort. The general movement is that younger speakers abandon the old literary readings, replacing them either with the colloquial reading or with the Mandarin pronunciation outright.

So the distinction is not only difficult, it is a moving target, and two speakers thirty years apart may genuinely differ. That is part of the larger story on the old and new Shanghainese page.

What a learner should actually do

Not very much, and that is the useful conclusion. You do not learn the doublets as a system; you learn words, and each word comes with its reading attached. Learn 物事 meh zy ‘stuff’ and 事物 zy veh ‘matters’ as two separate vocabulary items and the problem disappears.

What the knowledge buys you is the ability not to be thrown. When a character you know turns up sounding wrong, you now know there is a category for that and you can look it up rather than assume you misheard. Wiktionary labels the two readings explicitly in its Wu sections, which makes it the practical place to check. The wider sound system is on the pronunciation page, and why the writing gives none of this away is on the writing page.

Sources

1. Wikipedia: Literary and colloquial readings of Chinese characters — the mechanism, and the sources of literary readings in northern and southern Wu
2. Chinese Wikipedia: 文白异读 — Min most complex, Wu and Xiang intermediate, Beijing and Cantonese few; the Qian Nairong generational data on 文学 and 男生
3. Wikipedia: Northern Wu phonology
4. Wiktionary: 大 — Wugniu vernacular and literary readings
5. Wiktionary: 家 — literary and colloquial Wu readings
6. Wiktionary: 行 — literary and colloquial Wu readings
7. Wiktionary: 白 — literary and colloquial Wu readings
8. Wiktionary: 學 — colloquial in 上學, literary in 學習