Cantonese Romanization schemes

Question

The Cantonese romanization scheme that I know is Yale Romanization, which to me makes the most sense and is the easiest to read of any that I've seen. It seems to give a very good idea of how to pronounce things and is easier to read than, say, trying to remember which number goes with which tone. I'm aware, however, that there are others out there. Is Yale the most common and if not, what is? What schemes do you all use? Should I take the time to learn another?

Answer 1

This will give you a good background into each of the systems:

Cantonese romanization systems are based on the accent of Canton and Hong Kong, and have helped define the concept of Standard Cantonese. The major systems are Barnett–Chao, Meyer–Wempe, the Chinese government's Guangdong Romanization, Yale and Jyutping. While they do not differ greatly, Yale is the one most commonly seen in the west today.

I think it's all personal preference. As you mention the Yale system makes most sense to you, then I would stick with that.

Each system is about providing you a tool, but they all have the same aim. To enable you to understand Cantonese. To this end, there is no benefit in learning two of them thoroughly.

Another point is that romanization systems are limited in the benefit they provide for learning a language. They are covering a very minute amount of knowledge. So most students (in China) learn pinyin as a first step and then discard it. It is studied in the first month to become familiar with the sounds and letters and then it is only used for reading for a year or two after that. So don't rely on this as the cornerstone of your learning or as a regular tool for reading, use it as the most basic of tools, it is essentially just a 2 or 3 letter representation along with a tone to teach you pronounciation.

Answer 2

The two most common systems are Yale and Jyutping, the latter was invented as late as 1993. I think both are included alongside pinyin in Unicode's table of Chinese characters.

Most native speakers don't use and are not even aware of these systems, especially in mainland China where Cantonese speakers never study their native language at school - it's hardly considered a "real language". I have heard though that natives tend to use Jyutping, while Yale is used in nearly all English teaching material for Cantonese and is thus more popular among foreigners.

Then there is a third, obscure system with no tone information that nobody uses, except the HK government did when choosing all official, romanized place names, personal names, and so on. Oops. So neither system is good for reading romanized signs, maps et c!

I'd recommend learning and using Jyutping over Yale, for the following reaons:

• It's recommended by the Linguistic Society of Hong Kong, who invented it. (Not a huge reason, but noteworthy.)
• It is based on pinyin and uses the same letters for the same sounds. I think this is good for a Cantonese romanization system: many, many, many more people know pinyin, thus Cantonese becomes more accessible by being compatible with it and avoiding confusion.
• Jyutping distinguishes more sounds and is more accurate.
• Jyutping gets tones right, being straight-forward and consistent. It doesn't mix tone information and "spelling", like Yale (which sometimes uses double a:s or h:s to mark tones). It doesn't ignore tones entirely like the "government system". And there are only 6 tones, unlike the 9-tone model responsible for the linguistic urban myth of "look Cantonese has 9 tones, it's sooo hard to learn".

Yes, it's takes a bit of work to get used to those 6 tones, but tones aren't optional. It's much more frustrating when you eventually speak the language somewhat well, you're trying to read signs with unfamiliar Chinese characters and the romanized text totally lacks tone data! What are you supposed to do, read all permutations and ask a native which one makes sense?

• If you learn Jyutping you can use it to input characters especially in Linux-systems, but also in Windows and perhaps Mac OS X, I think, by using commercial software.