Last updated February 2, 2011

Tones and Jyutping

As mentioned elsewhere in these pages, I find the pronunciation and recognition of the tones in Chinese to be the most difficult aspect of learning.  Perhaps this is because I have a poor musical ear, it took me several months of playing guitar before I could hear the difference between an E and Em chord!

Unfortunately, the tones are absolutely critical if you are hoping to be understood.  You might think that you can rely on the listener being able to interpret your meaning from the context of the sentence but this is not always the case.  Many similar sounding words in Cantonese have conflicting meanings which is a great source of exasperation for students struggling with an already difficult language!  Consider, for example, 'buy' and 'sell'.  These are pronounced maai5 and maai6.  How this doesn't result in total confusion when trading is currently beyond me but I am living in hope that it will all become clear eventually.

Interestingly (or infuriatingly!), many words can have identical sounds and tones but mean one of several things depending on the context.  A quick glance at an online cantonese dictionary, shows that fo2 has characters that can mean fire, plenty, droplet, colleague and a numerical classifier for trees.  It is worth noting that other languages have homophones as well.  eg. in English, bow & bough, gilt & guilt, course & coarse, for & four etc.

Another problem is that many common words can mean some fairly offensive swear words if they are pronounced in a certain tone.  For example, there are two numbers between one and ten which can easily amuse or offend your audience if you are not careful. 

Of course, if you get several tones in your sentence wrong you will just end up speaking complete gibberish, don't worry, this is all part of the fun of learning the language!

The romanisation system for representing tones and pronunciation on this website is 'jyutping' (sometimes spelled 'jyutpin' or 'jyut ping').  It uses six numbers to represent the tones.  They are:

  • Tone 1: high level flat (or falling) - eg: faa1,bing1
  • Tone 2: rising to high level - eg: tou2, hoi2
  • Tone 3: mid level flat - eg: sai3, hei3
  • Tone 4: low level falling - eg: naam4, wu4
  • Tone 5: rising to mid level - eg: ngo5, jyu5
  • Tone 6: low level flat - eg: hai6,din6

Low, mid and high represent the tonal range of the speaker. So, speak tone 6 words as deep as you can whilst still sounding natural. Tone 3 words should be in your normal speaking voice and tone 1 words should be higher. Don't exaggerate these tones though, speaking tone 1 words as if you have inhaled helium isn't advisable! 

Throughout the site you will find audio samples of words and sentences containing all the different tones.

Please see Common Tone Mistakes for examples of the pitfalls in using the wrong tone!

If it is any comfort to people who are struggling with learning Cantonese, my current level of ability is to speak using only three tones consciously, normal, high and low!  In practise, I probably get more tones than this correct by unconsciously mimicking my teachers or people I've heard.   My speaking is pretty poor though...

This is an ongoing essay, so if you have any suggestions, corrections or advice, please contact me at cantonese@sheik.co.uk

CantoDict Tone Conventions

CantoDict uses the following convention to distinguish between literary and colloquial pronunciations. i.e. Literary are how text books say you should speak a word correctly, and colloquial is how people actually say it in real life. The latter is obviously most useful to a learner, but the former is also useful to know when looking words up etc.
  • mui6*2 : mui6 = literary; mui2 = colloquial
  • faa1 sang1 tong4*2 : faa1 sang1 tong4 = literary; faa1 sang1 tong2 = colloquial

As far as I am aware, we invented this convention for CantoDict, as normal jyutping does not show these distinctions.