Case Study
To look at tangible phonetic changes, I did a case study of Shanghai, a Wu dialect that is undergoing rapid change, using the speech of a friend's family. The two parents (in early 50s) and the child (early 20s) are native born Shanghai residents. They speak the Shanghai dialect at home. I used Section 2.4 (Internal Differences) of the Urban Shanghai Dialect Records as a guide.
[View Section 2.4] (in Adobe PDF format)
Section 2.4 of the Records lists some internal differences -- 9 features in the initials, 21 features in the finals, and 3 tonal structures -- in the dialect of urban Shanghai. These are based on a survey from the mid-1980s.
I checked all 30 features in initials and finals (but not tones) in the family and the results are summarized in the table below, using the same categories as in Section 2.4:
Initials
| Feature Index | Father | Mother | Child |
| 1 | 2 | 2 | 2 |
| 2 | 2 | 2 generally (but 户: v, 扶: β) | 2 |
| 3 | 2 | 2 | 2 |
| 4 | 3 (前谢: ʑ) | 3 (静谢: ʑ, 序: ɕ) | 3 |
| 5 | 1 | 1 (but 误梧: ɦ) | 1 (but 危: ∅, 梧魏: ɦ) |
| 6 | 1 | 1 | 1 |
| 7 | 2 | 1 (ʑ, dʑ usage mixed) | 1 (no dʑ at all) |
| 8 | 1 | 1 | 1 |
| 9 | partial unvoicing (e.g. 捕乎掉) | advanced unvoicing (e.g. 颂乎赠掉辅) | nearly complete unvoicing (父颂硕乎羡) |
Finals
| Feature Index | Father | Mother | Child |
| 1 | 2 | 1 | 1 |
| 2 | 3 | 4 | 3 |
| 3 | 1 | 2 | 1 |
| 4 | ᴇ uᴇ chosen based on character | ᴇ ø usage mixed even for same char. | ᴇ only |
| 5 | 1 | 1 | 1 (but iɪ very close to i) |
| 6 | 1 | 1 | 1 |
| 7 | 1 | 1 | 3 mostly (but 换: ɦuø) |
| 8 | 可: o; rest are ɔ | all are o; some -ɑ (Std. Chinese infl.) | 可: o; rest are ɔ and -ɑ |
| 9 | 1 (but 畅=唱: tsʰɑ̃) | 1 | 3 (no distinction) |
| 10 | 4 | 4 | 4 |
| 11 | 2 (mixed usage) | 2 (iŋ) | 2 |
| 12 | 2 (but 八 diff. from 滑) | 2 (all aʔ) | 2 |
| 13 | 1 | 1 (slight diff.) | 2 (though 达 diff. from 特) |
| 14 | 1 | 1 | 1 (but 却=曲) |
| 15 | 3 | 2 | 3 |
| 16 | 1 (but 月: ɦyɪʔ as in 2) | 1 | 1 |
| 17 | 1 | 2 | 2 for all (except 特 diff. from 夺) |
| 18 | 2 | 2 | 2 |
| 19 | 1 | 2 (retains a hint of the final uᴇʔ) | 3 |
| 20 | 1 (except 佛泼: oʔ) | 1 (only 佛: oʔ) | 1--> 2 (末泼佛: oʔ) |
| 21 | ɕyø | ɕyø | ɕyɪ (not commonly used?) |
Brief Discussion
The phonetic characteristics above fit well with the observations made in the Urban Shanghai Dialect Records. Specifically, the pronunciations of the parents correspond to the middle-aged stratum of the Shanghai dialect and the pronunciations of the child roughly correspond to the youth stratum of the Shanghai dialect.
The father's pronunciation retains features of Old Shanghai in more cases than the mother's pronunciation. Family history explains this to a small extent, as the father is a second-generation Shanghai native (grandparents on that side are from a different Wu speaking region) while the mother is at least a third-generation Shanghai native. Some Old Shanghai features are seen to be unstable. Some features have been unmistakably lost in the parents, for example: the implosive and bilabial fricative initials are lost (initials features 1 and 2); /tsi-/ /tsʰi-/ /si-/ /zi-/ have their initials merged into the aveolo-palatal ones (initials feature 4); distinction between /ɑʔ/ and /aʔ/, between /iəʔ/ and /iɪʔ/, and between /oʔ/ and /ɔʔ/ are lost (finals features 12, 15, 18). The consonantal losses in the initials are in line with Standard Chinese pronunciations. This may be a mode of change. On the other hand, the finals are lost in a different way, with the most aggressive losses occurring in glottal-stopped finals that are simply too similar and too short to fully articulate a significant difference.
Next, the child's speech takes on some, but not all of the newest features of the youth stratum of the Shanghai dialect as proposed in the Records. In the youth stratum, there has been a lot of influence from Standard Chinese -- and this is also showing its effects. The most unmistakable generational changes from the parents to the child are: the loss of the original /dz/ (initials feature 7), unvoicing of certain voiced (Middle Chinese) initials (initials feature 9), loss of distinction between /ɿ/ and /ʮ/, between /ã/ and /ɑ̃/, between /ᴀʔ/ and /əʔ/, and between /uoʔ/ and /uəʔ/ (finals features 1, 9, 13, 19). Some of the consonantal changes are continued progression from changes in the parents, but this is cut short by new changes induced by Standard Chinese, as in the case of unvoicing of initials (Standard Chinese has no voiced initials at all). The changes in the finals come three different ways: continuation of losses in parents (usually follows the less conservative parent when the two parents differ), loss of solitary distinctions (/ã/ and /ɑ̃/ are easily merged after /ãʔ/ and /ɑ̃ʔ/ are merged in parents, especially in light of Standard Chinese making no distinction between many such pairs), and once again, the loss of distinction in short glottal-stopped finals.
Conclusion
It seems from this case study that the Shanghai dialect is undergoing simplification in terms of number of distinguishable sounds in the initials and the finals. The current trend is in the unvoicing of voiced initials and the merging of, first, glottal-stopped finals, followed by the corresponding finals without the glottal stops. The primary influences appear to be from the speech of the parents and Standard Chinese.
After pointing out these observations to the friend, he said that some distinguishable sounds in his parents' speech were not even realized to be different by him. It would also be interesting to look at how the tonal contours and tone sandhi rules in the Shanghai dialect have been changing. They have been undergoing a certain degree of simplification over the years, also.
On to: References | Project Details