I've asked a similar question before:


but I want to make this more specific here.

I know not everyone agrees with tone accents in dialects/topolects - but just for the sake of argument, how should it be done?

I want to know how tone accents are dealt with in topolects/dialects when there are two falling tones.

Here's what I've found, so far:

Yale Cantonese seems to do it this way:

siː˥˧ = sì
siː˨˩ = sìh

So high falling tone = grave accent | low dropping tone grave accent + h on the end

while IPA does:


Hokkien seems to do it this way:

˧˩ to ˨˩ (31~21) = à
˧˨ʔ (32) = ah

Bigger drop gets a grave accent, light drop gets an h without an accent.

A simple standard seems to be:

Most obvious tone gets a grave accent - light drop gets an 'h'.

Here's my main question:

Is this a standard? Should all dialects follow suit?

I don't like that circumflex in the IPA one bit - seems quite confusing rather than anything else.

I'm inclined to use grave and double grave to denote the differences, myself - but I'd rather go a more "standard" route if possible.

  • 1
    I would advise you to be careful of your terminology. Accent has several meanings, but I think you mean orthographic accent here, which is better referred to as a diacritic. I was a little confused on first reading because I thought you were referring to some kind of pitch accent or to the interaction of tone and stress!
    – Michaelyus
    Feb 26, 2017 at 1:35

2 Answers 2


In essence, if I understand you correctly, you are asking for a standard for romanisation of tone diacritics. The straightforward answer of course is that each system has its own standard: Hanyu Pinyin for standard Mandarin is different to traditional Yale for Cantonese, which is different to Jyutping for Cantonese, which is the same as Sidney Lau for Cantonese.

With many learners whose first encounter with tone as a concept is with stnadard Mandarin Chinese through the orthography provided by Hanyu Pinyin, the diacritics are virtually featural; it is as if each diacritic was designed to represent the pitch of each tone as pronounced in the standard way. Neat, right?

This also works to an extent with Yale in Cantonese (including if you consider the high falling tone distinct from the high level tone, which is a minority in the 21st century). However, what to do with the low tones, which happen to be the same contours? Just add "h" after the vowel (or syllabic nasal consonant) and you're done! Not quite as featural as Hanyu Pinyin, but systematic enough.

However, Peh-oe-ji and its descendants are not very intuitive it comes to the representation of the tones in its orthography of Hokkien. If we look at the tones of Xiamen (Amoy) Hokkien, for which Peh-oe-ji was originally developed, through a strictly synchronic phonetic view, we get (in descending pitch order):

  • high level [including obstruent endings];
  • high falling;
  • (high) rising;
  • mid level;
  • low falling;
  • low with obstruent endings.

So we have two level and one kind-of level tones, one rising tone, and two falling tones. However, Peh-oe-ji was not developed with this view of the tone system; instead, it allocated the tones according to the historical four tone 平上去入 system of Middle Chinese, broken down by 阴阳 register, with 阴 going first then 阳. This is also technically the same view (let's call it the "historical" view) as the tone numbers for Cantonese in Yale and Jyutping.

The allocation of the diacritics seems fairly random, although an explanation from WH Medhurst's pioneering 1832 dictionary on (what would later be Malaysian) Hokkien can be found. Essentially, his tone marks are very subjective, introspective descriptions, and not directly related to pitch.

  • Do you happen to have a dictionary file distinguishing the high level and high falling tones in Cantonese? Jan 6, 2021 at 4:20
  • @kwaalaateimaa No; but according to CantoDict Forums, there are a few resources available.
    – Michaelyus
    Jan 6, 2021 at 12:14
  • I'm actually a previous commenter on that thread. Do you know of any other references? Jan 6, 2021 at 19:09

It seems to me that in each pair of tones you provided, the first of them is 去声 and the second of them is 入声. So they're two different tones rather than different ways of accentuation.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.