Shanghainese vs Suzhou, Ningbo and Hangzhou
These are Shanghainese’s actual relatives, and the relationships are not equal: one is the old prestige variety, one supplied words the city still uses daily, and one is only half Wu at all.
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.
The short answer
All four are Taihu Wu — Northern Wu — and the varieties of that group are mostly mutually intelligible with one another. A Shanghainese speaker can follow Suzhou speech with some effort and Ningbo speech with a little more. Hangzhou is the odd one: classified as Northern Wu but so heavily influenced by Mandarin that linguists have called it a hybrid, and some have called it a creole.
None of this resembles the gap to Mandarin, which is a different branch entirely and not intelligible at all. Nor the gap to Wenzhou, which is Wu and still unintelligible.
| Variety | Wu subgroup | Distance from Shanghainese |
|---|---|---|
| Suzhou 苏州话 | Suhujia — the same subgroup | closest; the old prestige variety |
| Ningbo 宁波话 | Yongjiang | close; supplied vocabulary to Shanghai |
| Hangzhou 杭州话 | Hangzhou, its own subgroup | Wu, but heavily Mandarinised |
| Wenzhou 温州话 | Oujiang — Southern Wu | not intelligible, despite being Wu |
| Mandarin | a different branch of Chinese | not intelligible |
An honest word about the intelligibility numbers
Several pages on this site quote a 2009 experiment by Tang Chaoju and Vincent van Heuven, which measured how well speakers of fifteen Chinese varieties understood one another and found that Beijing listeners caught 26% of words and 5% of sentences. It is worth being exact about what was tested, because it is the best data available and it is not quite about Shanghai.
The fifteen varieties were six Mandarin ones — Beijing, Jinan, Xi’an, Taiyuan, Chengdu and Wuhan — and nine southern ones: Suzhou, Wenzhou, Nanchang, Meixian, Xiamen, Fuzhou, Chaozhou, Changsha and Guangzhou. Wu was represented by Suzhou and Wenzhou. Shanghai was not in the study.
Suzhou is a reasonable stand-in, and this page is the reason why: Suzhou and Shanghai are in the same Suhujia subgroup of Taihu Wu, and Li Rong’s Atlas records the local varieties of that group as mostly mutually intelligible with each other. So a figure measured on Suzhou is a fair guide to Shanghai. But it is an inference from a neighbour rather than a measurement of Shanghainese, and every page here that uses it says so.
There is no published intelligibility experiment using Shanghainese as the stimulus variety that we have been able to find. If you know of one, the support link at the foot of this site reaches a person.
Suzhou: the variety Shanghainese grew up beside
Suzhou is the closest relative and the historical senior. Before Shanghai became the region’s great city, Suzhou was its cultural centre, and Suzhou Wu was the prestige form of the language — the variety in which opera was sung and in which the nineteenth-century novel 海上花列传 wrote its dialogue, even though the novel is set in Shanghai. The idiom about the softness of Wu speech, 吴侬软语, is about Suzhou above all.
For a learner the practical points are that the two are close enough that Suzhou material is usable with care, and that they are not the same: the tones differ, and so do a number of everyday words. Do not learn from Suzhou audio and expect to sound like Shanghai.
Ningbo: the variety that lent Shanghai its words
Ningbo is in the Yongjiang subgroup, about 4.5 million speakers, and its influence on Shanghainese is out of all proportion to its size — because Ningbo people moved to Shanghai in very large numbers as the city grew.
The clearest trace is the word Shanghai is known by across China. 阿拉 a la ‘we, us’ — the Shanghainese for we, and the most recognisable word in the language — came in from Ningbo and displaced the older Shanghai form 我伲 ngu nyi ‘we’, which older speakers still use. The full story is on the pronouns page.
That is a useful thing to know about Shanghainese generally: it is a city language assembled over about a century and a half of immigration, and some of what feels most characteristically Shanghai arrived from somewhere else.
Hangzhou: Wu with a Mandarin accent
Hangzhounese is the interesting case. It is traditionally classified as Northern Wu and it is given a subgroup of its own, with around 1.9 million speakers in the urban core. But it has taken so much from Mandarin that it is hard to place: phonologically it lacks features that Suzhou and Shanghai share, while also doing things no Mandarin variety does. Linguists have described it as a Mandarinicised Wu hybrid, and some have gone as far as calling it a creole.
The reason is historical. When the Southern Song moved the imperial court to Lin’an — today’s Hangzhou — a large northern-speaking population came with it, and the prestige reading layer that court brought is still detectable across Wu. Hangzhou got it at full strength.
So of the three, Hangzhou is the one where a Shanghainese speaker notices something structurally different rather than just a different accent. It is also the one whose local identity is least tied to its speech: Hangzhounese matters to people from Hangzhou, but not the way Shanghainese matters in Shanghai.
So which Wu variety should you learn?
Shanghainese, unless you have a specific reason not to, and the reason is not linguistic. It has the most speakers of any single Wu variety, by far the most material, the only real body of recorded audio, and the only television series a learner can work with. Suzhou has the prestige and the opera; Ningbo and Hangzhou have almost nothing written for outsiders.
And the effort transfers. Because Taihu varieties are largely mutually intelligible, learning Shanghainese gives you a great deal of Suzhou and Ningbo for free, in a way that learning Wenzhounese would not. What exists for Shanghainese is collected on the resources page.
Sources
- Wikipedia: Wu Chinese — Taihu subgroups, speaker figures and mutual intelligibility (citing Li Rong, Language Atlas of China, 2012)
- Tang, C. and van Heuven, V. J. (2009). Mutual intelligibility of Chinese dialects experimentally tested. Lingua 119(5), 709–732.
- Reed College Sociolinguistic Artifacts: summary of Tang and van Heuven (2009), listing the fifteen varieties tested
- Wikipedia: Hangzhou dialect — classification, Mandarin influence and speaker figures (citing VanNess Simmons 1999; Li 2012)
- Xu, Baohua and Tang, Zhenzhu (1988). 《上海市区方言志》, p. 418, on 阿拉 entering from Ningbo — cited in Wikipedia: Shanghainese.
- Wikipedia (Chinese): 海上花列传
- Wikipedia: Suzhou dialect
- Wikipedia: Ningbo dialect