top of page

My Blog (WordPress)

Follow me on Tumblr        or WordPress        or find my related stories on Medium 

  • Tumblr Social Icon

Keith Tse


  • Writer's pictureKeith Tse

Cantonese phonology reprised

Now let’s get serious. I mentioned before that there were numerous word-initial consonants in Cantonese that correlated with word-initial consonants in other dialects (e.g. Mandarin) and that these correlations showed us the original distribution of these consonants in proto-Chinese before their widespread loss in modern Cantonese (especially HK Cantonese). The result of this is that there is now a rather large amount of homophonous ambiguity between previously distinguishable words e.g. [oi] vs [ngoi], or [ngaam]. While those who are linguistically trained in comparative-historical linguistics can reverse the comparative method and reconstruct the original word-initial phonemes with relatively ease, this is impossible for most speakers of modern Cantonese who are linguistically untrained (and disinterested in all things apart from material good). There hence needs to be some strategy for disambiguation, since it would not do to have so many homophonous lexical items that would just impede efficient communication. There are some alternative disambiguating strategies that are really quite interesting. Let’s take an example: 安 /on/. For those of you who have learnt Chinese hieroglyphics before, you will be aware that Chinese character-writing is significantly (though not universally or even regularly) conditioned by orthographic radicals which form either conceptually or phonologically dissectible units for character-formation. The character 安 is one such radical and frequently recurs in numerous characters as a phonological (not conceptual) unit e.g. 案/按/鞍/垵/胺 and many others. But since all of these characters are homophonous, there needs to be a way to set them apart, even if none of these are homonymous (as mentioned, the character 安, which means generally ‘tranquility’, is not a semantically coherent unit and does not apply semantically or conceptually to any of its orthographic cognates). One such way is to insert a hyper-correct word-initial consonant i.e. one which was never there in the first place e.g. 案, which is pronounced as [ngon] in modern Cantonese even when the Mandarin counterpart is just [an]. This hyper-correct pronunciation is 案 [ngon] is very robust in modern Cantonese, since even the younger generation consistently pronounce it with an initial consonant [ngon], which indicates that pronunciation can be conditioned by non-phonological considerations. One might also add that the meaning of 案 ‘criminal case/investigation’ is somewhat strong and serious and the hyper-insertion of a word-initial phoneme [ngon] supports its basic semantics. Never knew that our tongues could do such wondrous things, especially since I am a native speaker of Cantonese and have never had to think hard about these things, unlike English and Romance and all other foreign languages which I have had a tough time getting my head around (and still do…).

3 views0 comments


bottom of page