I’ve been suffering insomnia for the last three days, a combination of mosquitoes having acquired stealth technology and somatoform cockroaches. I heard if they climb into your ear canal they can’t back out. The real ones, that is. Not the mindroaches.
This is a quick mockup of my latest insomnia-induced idea. In the image on the left, the first number in each two digit sequence denotes the beginning of the tone; The second denotes the end tone, just as has been used throughout the multitude of sources marking tone.
The idea behind this is that one (me, I guess) would be able to map out the changes of tone curves from one town to the next throughout the Delta. They’re colours instead of numbers so that they can be arranged on an actual map to give a better sense of the changes at a glance, thus the second image. Let’s say this represents 阴平.
A similar map would be available for each tone. It would of course be less grid-based and more organic based on the information available for each dialect. Right now I’ve got tone curves for about 70 dialects. It’s possible to go from one town to the next and see how things change based on the number in any two town’s curves, but it would take freaking forever and wouldn’t provide much information on the general trends.
So that’s my idea. I’m posting this here to get responses and to hash out some of the possible difficulties. One such difficulty is dipping tone curves. With three numbers, you’d need another layer of information to show the value. Colour would work well with a shift in hue but that’s asking a lot of the viewer’s ability to distinguish subtle changes in hue to mark a third tone. In other words, it wouldn’t work. Saturation variations could work too but it runs into the same problem as with hues.
That is unless you required less accuracy in the middle value, so that ˦˩˧ and ˦˨˧ would display the same. That is, dipping tones may get a bit of red to them and tones that ‘bulge’ get some grey.
I hope that was clear. Thoughts?












Cockroaches can crawl into your ears. It happened to my wife one fine morning at 5 AM here in NYC; the noise of the roach moving about in the ear canal was unbearable, and we had to take her to the emergency room at the Eye and Ear Hospital. Fortunately, the treatment’s simple and easy: flush the canal with mineral oil. The roach dies almost at once, and you then just tip your head over and let the canal drain. So get yourself some mineral oil and an eyedropper, just in case.
The map: unutterably cool and brilliant idea. Go for it. (It would be wondrous to see such a map for the whole of Han-speaking China, but that’s probably beyond anyone’s resources today.) I look forward to looking at maps for each tone with wonderment and awe.
Dipping tones: I don’t know anything about these in Wu, but my sense in Mandarin, FWIW, is that it’s the rising part of the tone that counts. If you pronounce Mandarin tone 3 as just low-rising, you aren’t misunderstood; intuitively I think of tone 3 as “a little fall, a lot of rise”, like a (right-handed) check mark.
Good god. Now I know what I’m doing today: looking for mineral oil.
Treating the dipping tones as simple rising tones could also work to get a general sense of things, though I think it may be worth keeping the dipping nature represented. If, say, Zhejiang Pinghu has the Shang tone dip, but then in nearby Huzhou it’s just rising, that may be something I’d like to see.
Another option would be to make the map and then just overlay a little D where they dip, if dipping turns out to be of little importance.
Very cool. Although, I must admit, I’m not sure exactly what is being mapped. By tone changes, do you mean variation from 普通话? Or between the dialects respectively?
Cool nonetheless though. You’ve got real knack for this stuff.
Hey thanks.
It would be variation from one 方言 to the next. Maybe for example in Kunshan a tone is high to low but in Suzhou the same one is mid to low and Wuxi has mid dipping to low but then back to mid. On the map then it would fade from a blue to a green, Kunshan to Wuxi. Each square/colour in the first image represents one tone curve.
You could alter the shapes of the boxes for dipping/cresting. (For example, bulging vs squeezing shaped squares.)