Linguisteria e Programmettas

Cantonese–Mandarin sound correspondences

Intro

As part of my Cantonese self-studies, I have decided to compare the sounds of Mandarin and Cantonese, which are closely related to each other. I am a native Mandarin speaker, and as I learn more Cantonese vocabulary, I begin to notice recurring patterns in the pronunciation of characters, in which similar sounding characters in Mandarin would result in those characters sounding similar in Cantonese as well. Thus, I conducted this experiment to confirm my intuitions.

This page is subject to change; when it does, I will create a changelog detailing the changes.

Datasets

I collected 3500 characters from the "Table of General Standard Chinese Characters", which consist of the most commonly used characters (and thus would exist in both Mandarin and Cantonese). One source of issue is that some characters, such as 行, are polyphonic, and thus cannot be easily mapped between the two languages. For simplicity, I have removed them from the dataset, and used monophonic characters exclusively.

Format

I separately compared the initials (the first consonant of a syllable) and the finals (the portion of a syllable after its first consonant) of characters in Mandarin and Cantonese. For Cantonese, I removed the ending consonant (-p, -t, and -k), since Mandarin does not have them and would excessively fragment the results (since if I included the ending consonant, one Mandarin final would correspond to three different Cantonese finals).

In addition, w- and y- are considered a part of the final in Mandarin, but they are classified as initials in Cantonese. In my comparison, I treat them as initials in Mandarin when they are not preceded by another consonant, and as a part of the final otherwise.

For Mandarin syllables, this page uses Pinyin; for Cantonese, this page uses Jyutping.

Initials

The following graph shows the frequencies of Cantonese initials that correspond to the selected Mandarin initial.

Select a Mandarin initial:

Observations

For the most part, the initials map between the two languages intuitively; Mandarin initials correspond to their idential counterparts in Cantonese. However, the results also have some notable exceptions:

Finals

The following graph shows the correspondences between Mandarin and Cantonese finals. In the dropdown, the underscone ("_") represents the syllabic consonants (zhi, chi, shi, zi, ci, si, ri in Pinyin). This distinguish them from the /i/ vowel, which appears in syllables like "ti".

Select a Mandarin final:

Observations

Cantonese has more finals than Mandarin does, so a single Mandarin final will often map to many Cantonese finals. Moreover, Cantonese finals can also end with the consonant "-m", which is not possible in Mandarin. Here are some notable comparisons:

Tones

Perhaps not surprisingly, the tones in Mandarin and Cantonese have correspondences as well. The following graph shows the mapping between them. Click on "Invert graph" to flip between the datasets.

Invert graph

As the graph indicates, the contours of the tones do not correspond directly (i.e. a rising tone in one language does not translate to a rising tone in the other), but the relationship is still clear, going from Mandarin to Cantonese:

Exceptions do exist though; since the checked syllables in Cantonese (those that end with -p, -t, or -k) will always have a flat tone, characters that would have had a falling or rising tone (as predicted by the patterns listed above) would instead have a flat tone. As an example, 十 has a rising tone (tone 2) in Mandarin, but instead of a falling tone, due to having a checked syllable that ends with -p, it instead has the low flat tone (tone 6).

On the other hand, mapping from Cantonese to Mandarin is fairly straightforward:

Of course, in any situations, exceptions exist.

Acknowledgements

Source code

The source code for computing the results on this page can be found here.