Wu Dialect Ain’t Nothing to F*** With— or, How Learning Hindi Made Me Better At My Mother Tongue

Kevin Sun
Sun Language Theories
9 min readAug 2, 2017

--

Population density map of the northern part of the Wu dialect area. (source: European Commission Joint Research Center)

“Native language,” “first language” and “mother tongue” aren’t always straightforward concepts. Depending on how you interpret these terms, I might already be straining the definition by claiming to have two of them — Mandarin and English. Well, what if I told you I might actually have a third native language?

That potential third native language would be Shanghainese, the most widely-spoken sub-dialect of the Wu dialect of Chinese. (As usual for Chinese, “dialect” here actually means “a separate language” and I’m using “sub-dialect” to mean a dialect of a “dialect.”)

There are some good reasons that I might consider Shanghainese my mother tongue — first of all, it is literally the first language of both of my parents. I grew up hearing it spoken all the time, especially in family settings among people of my parents’ and grandparents’ generations, and could understand it quite well.

But there’s also a pretty good reason to not consider it my native language: I never spoke it very well growing up.

That said, I have gotten better at Shanghainese lately, though this improvement came about in a wildly roundabout manner with an unlikely detour—specifically, learning Hindi was one the key things that helped me improve my Shanghainese pronunciation a lot.

Last week I wrote a bit about the world’s top ten native languages. Let’s see which languages just barely missed out:

The field gets a lot tighter after the top 10 — wouldn’t be surprised if the rankings have been shuffled a bit since 2010. (source: Wikipedia)

Check it out: Wu has more (native) speakers than French!

It also has more speakers than Cantonese, for that matter. Oftentimes, when I tell people I speak “Chinese,” they’ll kindly check my Mandarin privilege by asking me to clarify if I mean Mandarin or Cantonese. Which is great, and it’s cool that lots of people are aware of these nuances these days, but… Chinese has more than just those two dialects.

Of course, there’s a reason Cantonese is more well known than Shanghainese. The overseas Chinese diaspora was historically heavily Cantonese-speaking (especially beyond Southeast Asia, where Hokkien and Hakka are also prevalent), so people in the West would have had more contact with it. Furthermore, Cantonese has significantly more official recognition and prestige than any other non-Mandarin dialect, as an official language of Hong Kong and the language of a substantial music and film industry.

Wu has much less of an official presence in the public sphere. I can remember one well-known Shanghainese sitcom that made heavy use of dialect, and traditional theater forms from Shanghai and the surrounding area still use regional language varieties, but the vast majority of modern pop culture in the Wu dialect area is now produced in Mandarin.

Probably most significantly, we were not supposed to speak Shanghainese in school — although many of my classmates would goof around in Shanghainese after class, and even teachers would sometimes slip into dialect when they got mad. Since I didn’t start living in Shanghai until I was ten, I never spoke much Shanghainese at all, and when I did, I did so with a noticeable accent.

One major issue with my Shanghainese pronunciation was my inability to distinguish between voiced and unaspirated stops.

Okay. So… what are those?

Time for a quick phonology primer — if you know this stuff you can skip to the Venn diagram below.

English has two sounds where you block and release the flow of air with your tongue at the back of your mouth— “k” and “g” — which are known as “velar stops.” Chinese also has two velar stops, which are also written in Pinyin as “k” and “g.” Spanish has two velar stops as well (and French, Portuguese, Italian… etc. etc. I’ll use Spanish as the example here since it’s probably most well known).

But here’s the catch! None of these languages use the same two velar stops:

  • in English, “k” comes with a puff of air after it (i.e. it’s “aspirated”), and “g” is pronounced with vibration of the vocal cords (so it’s “voiced”)
  • Mandarin “k” is aspirated like in English, but Mandarin “g” is not voiced and not aspirated either
  • Spanish “k” is unvoiced and unaspirated (like Mandarin “g”!) while Spanish “g” is voiced like English “g.” (English-speaking Spanish-learners often pronounce Spanish “k” in the aspirated English way, which is one of the major tells of an Anglo accent.)

Or, in Venn diagram form:

Wonder what the middle section could be 🤔…

The same relationship applies for other positions in the mouth as well, such as “p” and “b” (labial stops) or “t” and “d” (dental stops).

Before I knew anything about linguistics, I grew up assuming that English “g” and Mandarin “g” were exactly the same — I mean, why else would they be written using the same letter? But if I had actually pronounced them exactly the same way, that would mean that my Mandarin pronunciation was technically slightly off.

That might be confusing enough, but get this: Shanghainese has all three of these velar stops! So we’d complete our chart like so:

… Oh.

Despite having listened to Shanghainese regularly for most of my life, I didn’t realize this basic phonological fact about the language until I was in high school, once I started learning a bit about phonology while studying other languages.

For example, I had grown up believing that the word 讲 (to speak) and the word 戇 (stupid) were pronounced exactly the same way in Shanghainese (roughly like “gong”, but with a French-style nasal vowel). In fact, the former is pronounced with an unvoiced Mandarin “g” while the latter is pronounced with a voiced English “g.” My inability to distinguish “speaking” and “stupid” resulted in me speaking stupidly. Sometimes when I tried to speak Shanghainese as a kid, my classmates would tell me it sounded wrong, but they didn’t know how to explain why exactly it was wrong — because they didn’t have the technical terminology to do so.

(This three-way distinction in Wu dialects is a holdover from ancient Chinese. Mandarin, Cantonese and Hakka have lost their voiced stops, while Wu and Min (which includes Hokkien/Taiwanese) still maintain the distinction.)

If this was just an issue with my Shanghainese pronunciation, I probably wouldn’t have minded much, but this phonological confusion had significant implications for my language learning more generally. Most Indian languages (especially the northern ones) require you to distinguish between all of these sounds. If I couldn’t even figure this out in a language I’d been listening to since birth, how would I ever be able to distinguish these sounds in a foreign language?

This was one reason I ended up putting off Hindi for so long — I honestly wasn’t sure I’d be physically capable of speaking it correctly. (Supposedly, one of the major challenges in learning a language as an adult is learning to distinguish sounds that you consider the same, which is much harder than simply learning to produce a completely new sound.)

What’s worse, Hindi’s sound system is even more complicated than Shanghainese — for each mouth position where English has two stops and Shanghainese has three, Hindi has FOUR. The fourth one is voiced and aspirated at the same time.

For example:

  • the word for India itself is “Bhārat” — the vocal cords vibrate and a puff of air comes out at the end of the first consonant
  • the word for “attention” is “dhyān” (which was borrowed into Chinese and Japanese and turned into “zen,” by the way)
  • “Gandhi” also has the same sound — that’s why it’s spelled with an “h”

These were also tricky for me—when I first got started, I would frequently mishear them as their unvoiced equivalents (e.g. “Gandhi” as “Ganthi,” “Bhārat” as “Phārat”).

Indo-Aryan velar stops and their relationship to other languages. (See a full alphabet chart from Omniglot here.)

The phonological challenges of Hindi don’t end there (for example, “t”s and “d”s come in “retroflex” and “dental” versions), but this voicing and aspiration thing is a major key.

Once I finally did start studying Hindi, this was one aspect of the language I made sure to focus on getting right. After several months of repetition and careful listening — as well as just learning more vocabulary to help deduce the correct pronunciation from context — I think I learned to sort out the distinctions in pronunciation pretty well.

Another distinguishing feature of Wu is that its word for “shit” is different from that used in all other dialects of Chinese. 💩 (source: z3z4)

Of course, my reason for learning Hindi wasn’t to get better at Shanghainese — it was to learn Hindi. So I didn’t immediately start putting my newfound phonological discernment to use elsewhere. In fact, the last time I visited Shanghai was before I had studied any Hindi at all.

I eventually did get to test out my leveled-up Shanghainese phonological skills during a random encounter in New York. I was at a language exchange event one weekend, and after practicing a bunch of Spanish and French and Russian and whatnot, I ran into someone from Shanghai who was there to practice English. It had been a long time since I’d even attempted to speak Shanghainese, but I was very much in the language-practicing mindset at the moment, so I decided to give it a shot right there. It was probably the most successful Shanghainese conversation I’d ever had, and I’ve had a couple more since then. Next time I go back to Shanghai is going to interesting.

(The phonological sensitivities of Hindi weren’t the only thing that helped here. Weird Germanic vowels, the Ukrainian “g” sound, tone-awareness I had picked up from Cantonese and Lingala, plus etymological knowledge I had gleaned from Japanese all helped me to speak Shanghainese with much more precise pronunciation than before.)

Sub-dialect map of the Wu dialect area (source: baike.com)

A thousand years ago, much of modern Shanghai’s city center was still under water. The land gradually rose out of sea with the piling up of sediment carried downstream by the Yangtze River.

And two hundred years ago, Shanghai was still a small fishing village, until it was opened to foreign trade after the First Opium War. Shanghai’s dialect, like the land it was built on, emerged out of nothing as other Wu dialects piled up and mixed with each other, brought in by an influx of migrants from nearby areas — with Suzhou and Ningbo influences being most prevalent. For this reason, Shanghainese is sometimes considered the koiné language of the northern Wu region.

As I mentioned in my last article, dialect continuums are one of my favorite things in linguistics, whether it’s in South Slavic or Arabic or Punjabi. So I guess it was only matter of time before I started investigating the Wu dialect area. I haven’t gotten my hands on any academic books on the subject (yet) so I’ve been mostly playing it by ear — just listening to samples of various dialects to see what the differences are. The internet is great for this — I’ve come across a series of folktales in Suzhou dialect, nursery rhymes in Hangzhou dialect, and children’s books in Ningbo dialect, and it’s been entertaining to hear these language varieties which I can still mostly understand, but just seem slightly off.

(There’s lots more listening material in Shanghai dialect itself, too. For example, here’s the full biography of Kuomintang spy boss Dai Li in the form of a Shanghainese audiobook.)

The way learning Hindi helped me get better at Shanghainese is just one example of the strange ways that language learning can compound on itself. The benefits are obvious when you’re studying languages that are closely related to each other, but even seemingly distant languages will often have some kind of unexpected overlap.

--

--