Or at least why they may be much less important than in Mandarin.
I’ve argued that tone is important in Wu, even if focusing just on a dialect with a drastically limited tone set such as the Shanghai dialect. While most Wu dialects have 7 or 8 tones (including those immediately outside of Shanghai), Shanghainese has been reduced to about five. Maybe less since two of those 5, the 阴入 and 阳入, are mostly just reduced form of the 平声 and what’s called the 舒声 which is really just an amalgamation of all the non-ru 入 in the lower register (阳). Many people, locals included, will tell you that after the first word of a sentence the rest of the words don’t matter. I wouldn’t go this far but I could see taking the first and last into account and then dropping everything else to mid-level tones and still getting by just fine.
Previously, I’ve written that I think the reason most people neglect tones or feel they can be neglected in Shanghainese is that they just never get far enough into the language to really see how much they matter. You can have an atrocious “厕所在哪儿” in Mandarin and get by just fine1.
But I think I may have been wrong in taking that position. I still think there’s something to it, at least in the early stages of learning the language. I do think you can get by for a good while with crap tones in Shanghainese as long as your pronunciation is decent (mine’s not, for the same reason I could never say “não” right when studying 巴西话 years back). The following is from the Wikipedia article on tonogenesis, emphasis added.
Very often, tone arises as an effect of the loss or merger of consonants. … In a non-tonal language, voiced consonants commonly cause following vowels to be pronounced at a lower pitch than other consonants do. This is usually a minor phonetic detail of voicing. However, if consonant voicing is subsequently lost, that incidental pitch difference may be left over to carry the distinction that the voicing had carried, and thus becomes meaningful (phonemic).
So Mandarin without voiced consonants has tones and they matter a great deal. Wu, or specifically Shanghai Wu, has voiced consonants in addition to the aspirated and un-aspirated consonants and tones matter a lot less. If I remember correctly, there are only 412 different syllables in Mandarin. Wu not only has initials not found in Mandarin, but finals as well at least in the form of [ʔ]. Granted, in the case of [ʔ] it always takes 入声, but then maybe that’s the point. As long as that stop is evident in speech, you wouldn’t really need to hit the tone right. Again from the Wikipedia article:
In the Moka texts3 we have the same deal. The tones were not written in those books (unfortunately) but the words which would be pronounced with the entering tone are marked in their spelling. One of the characters in the story 《ㄐㄧ ㄉㄢ ㄍㄠ ㄌナ ㄌㄧ ㄑㄧ ㄗオ》 is named ‘Li-tok (ㄌㄧ-ㄉㅏ), the letter k in “-ok” (here ㅏ) marks the stop [ʔ] meaning we can tell that syllable was either 阴入 or 阳入.
With tones tied to consonant voicing in this way it would be one more point for the “tones are irrelevant in Wu” people. Or at least, the “mostly irrelevant in Wu” people.
- – -
1. I know because I did just that when I first got to China. At fast food joints I said “一号套餐” with 号 as 好 for the first six months and never once had it matter. Now in a grad program the distinction is rather important and with much of the context stripped by the miserly use of words that is 文言文, tone matters a whole lot more.
2. Worth noting for Wu, in Middle Chinese 入声 was taken by syllables ending in [p], [t] and [k], now a glottal stop [ʔ] in most modern Wu dialects. You can see this in the transcription in Wang Ping’s book on Suzhou dialect, the Moka Mission texts of the 20s and the Shanghaihua Da Cidian, pinyin edition.
3. The bopomofo used in the Moka texts was a modified form of the bopomofo used today. For the additional letters, see this old post. I’m using ㅏ here and ト there but they’re actually the same letter. The typeface used in the original text makes it unclear what the form really ought to be, and I have yet to find other texts using the symbol that are not from the Moka Mission.
I can understand Punjabi to a decent extent (having grown up surrounded by Punjabi speakers) and I still find it hard to believe that it has phonemic tones. All the native Punjabi speakers I know know nothing about tones, or for that matter, even have any vocabulary in the language (words like 声调 or 入声字) for talking about them.
If Wu is somewhat similar to the example of Punjabi, then I can see why people would say tones are unimportant. However, Wu still has loads more tones than Punjabi (I think Punjabi has only a low level and high level tone) and so the distinction is still important to maintain.
People know surprisingly little about the languages they speak. And they’re very trusting of the way it’s taught to them when young. Not being taught that there are tones pretty much makes people oblivious to them being there. Try telling most Chinese people there are 5 tones in Mandarin (the fifth being the neutral tone). They’ll never believe you, because that’s not how tones are discussed.
I do still thinks tones are incredibly important to Wu. It’s just interesting to see a link between voicing and tones shedding some light on a possible reason why it’s deemed less relevant.
If I correctly understand the discussion I’ve read of Shanghai sandhi (from Chen’s _Tone Sandhi_), the tone of the first syllable of each of the content words matter.
And in the other dialects that don’t have as much merger between the tone classes or don’t have checked syllables, the voicing is a smaller part of the story.
That’s true for the most part. Though it seems it’s not the case that the following syllables don’t matter. It’s more that they follow a predictable pattern from that, and depending on that first syllable the second might be mid or it might be high but you can’t just pass them all off as middle tone after the first and call it good.