November 26, 2007

Lazy Pronunciations

Lazy pronunciations (or lazy sounds / 懶音) are the "left-over" phonetic traits from childhood for some native-speakers. Some sounds are difficult to pronounce for a child and they are replaced by similar sounds which would take less effort to pronounce (hence, lazy).

As the child grows older and her linguistic ability has developed better, the lazy-sounds will be replaced by the correct sounds. The correction sometimes happens naturally (under peer-group pressure perhaps), and sometimes requires guidance (or force) from the adults.

Unfortunately, this process might not happen to some people, for many different reasons. This is why when people speak with lazy sounds, they sound like a child.

Common categories of lazy sounds include:

  1. 'n' initial replaced by 'l'.
    E.g., 男 (man) - naam4 becomes laam4
  2. 'gw' and 'kw' initials reduced to 'g' and 'k'.
    E.g., 過 (pass) - gwo3 becomes go3
  3. 'ng' initial disappears.
    E.g. 我 (I) - ngo5 becomes o5
  4. 'ng' final replaced by 'n'.
    E.g. 講 (talk) - gong2 becomes gon2
  5. 'k' and 't finals confusion.
    E.g. 塞 (block) - sak1 becomes sat1
  6. 'ng' final replaced by 'm'.
    E.g. 五 (five) - ng5 becomes m5
In recent years, many young pop-stars suffer seriously from lazy pronunciations and this seems alright, if not trendy. As a result, many youngsters follow them. This was certainly not the case in the past, when lazy pronunciations were less tolerated. I still remember when I was a kid, a children TV program was launched for only a week or so, and one of the hosts got replaced. The reason? Many parents complained that she spoke with lazy sounds.

Here is a video clip of Professor Au-Yeung (a.k.a the handsome professor) appearing in a Cantonese language forum. He demonstrates throughout his speech the lazy pronunciation of the word 我 (I) in which the 'ng' initial is eliminated. One noticeable instance is at 1:07 (-4:38) minute .


GSJ said...

There is nothing lazy about this; these are simply perfectly reasonable sound changes that the Cantonese language is undergoing. Your explanation is a combination of folk linguistics and pseudoscience. For instance, here's a study on childhood phonological development:

As you can see, syllable-initial /n/ usually appears around nine months, whereas /l/ does not appear until after the two-year mark.

Edwin said...


The cause of lazy pronunciations is a controversial topic, and I thank you for your time in leaving me a comment.

The study you have referenced only addresses the English speakers. A similar study for Cantonese can be found here.

The above study does confirm the confusion of some sounds listed in my post (e.g. 'kw' initial and 'k' final). As for sounds like the 'ng' initial removal, I am a first-person witness of seeing children doing it because it is easier to pronounce. Call my experience 'folk linguistics' or 'pseudoscience' if you like. I prefer to call it 'empirical evidence'.

As for the 'n' and 'l' sounds, how early an individual sound is developed in the childhood does not necessarily reflect how easy it is pronounced in a sentence as a whole. It also depends on the language itself. In the case of Cantonese, many native-speakers can testify that 'lei5' simply takes less effort to pronounce compared with 'nei5' when used in a sentence.

Why do the children choose those sounds in the first place? Some say it is the influence of the public media. I would say it is a bit of that and also in some cases, the sounds take less effort to pronounce. Then the major blow is the lack of awareness of the adults in correcting the lazy sounds produced by the children.

It is one of my objectives of this series to fight against the myth many people believe that phenomenon of lazy pronunciations is part of the Cantonese language evolution. I will have more upcoming posts on this.

I would appreciate in your future comments, if your could append phrases such as "I think" or "in my opinion" to any subjective comment on other people's opinions if it contains seemingly provocative words.

Once again I thank you for your opinion.

GSJ said...

Language change is always controversial, because speakers of a language are extremely sensitive to differences in how it is spoken, and have very visceral reactions to these differences. There is, therefore, a universal tendency to assign derogatory labels to people who talk differently ("lazy," "uneducated," and so forth).

However, none of these explanations are actually justified. Matthews & Yip (1994; Section 1.5) talk about how Cantonese pronunciation variation (such as the n/l variation) is "quite systematic, following variables such as age, class and gender," and that "this distribution in many cases clearly indicates sound change in process." (For more on sound change, check out Lyle Campbell's "Historical Linguistics," second edition. For Cantonese in particular, I recommend "Modern Cantonese Phonology," by Robert Stuart Bauer.)

These "folk" explanations are also subjective (and thus cannot be considered objective and empirical, but only anecdotal); those native speakers you cited find it easier to pronounce the /l/ because that's how they talk. "Lazy pronunciations" are actually "default pronunciations," which is why deviating from them always requires effort. It would take a lot of conscious effort for an older Cantonese speaker to talk exactly like a young person for an extended period of time without slipping up and "lazily" saying nei5 or ngo5 every so often.

In fact, I can personally attest to this: I'm learning Cantonese through Pimsleur lessons, and the speakers on those tapes use the old-fashioned nei5 for "you," so this is what I've grown used to. I therefore have to make an effort to say lei5 instead of nei5; for me, the /n/ is the "lazy pronunciation!" And of course, it's not because I find one sound easier to pronounce than the other; when I speak English, I don't replace /l/ with /n/ or vice versa, and to do so would take effort.

You also said that you are a first-person witness of children omitting initial ng-. I'm sure that you've witnessed this, but I question your explanation of what causes it. Just because children are doing it doesn't mean that it's a defective pronunciation. The fact of the matter is that the vast, vast majority of children learn to talk exactly like their peers (and NOT their parents, grandparents, or teachers). I guarantee you that if you went to a Cantonese-speaking elementary school, all the children would be addressing each other as lei5; they can't all be lazy.

Also, you pointed out that the document I cited was a study of English speakers; this is true, but it is my understanding that children pick up speech sounds in a certain order, and that the language they speak doesn't have much effect on which sounds develop first. For instance, the nasals /n/ and /m/ always develop very early, and as it happens, almost every language in the world has at least one nasal phoneme. (The early development of /m/ leads many people to believe that this is one explanation for why terms like "mama" for "mother" seem so common cross-linguistically.)

Anonymous said...

Well this is definitely perplexing when it extends to the same slippages in foreign words.

I have a colleague for whom Bangkok will remain Ban-cot, although, strangely, she is perfectly comfortable with 'k' endings in Cantonese.

Michael said...

I would like to expand upon a couple points here.

(1) Regarding the notion of 'lazy' sounds, a common misconception is that /lei5/ take less physical effort than /nei5/ (or substitute any n/l word). This, of course, is not true, as any phonetician will attest to. Furthermore, the actual difference in muscular effort between these two sounds is so slight that no one could possibly be sensitive to it. It somewhat like being able to detect the additional weight of a paper clip attached to a 5 kg bar bell. As gsj points out, the 'difficulty' comes in consciously remember to alter one's typical speech. The actual n/l distinction is arbitrary.

(2) No one has brought up, yet, the topic of hypercorrections. When people try to correct "lazy speech", they often unwittingly introduce yet newer sound changes. A rather common one concerns word-intial /ng/-deletion, as in 我 (ngo5 --> o5). In an attempt to retain the word-initial /ng/, this sound is sometimes introduced where it never originally occurred. 爱, for example, was originally, /oj3/, but is now often pronounced /ngoj3/. None of characters with upper register tones (tones 1-3) ever originally carried word-initial /ng/, although as a learner of Cantonese, I've had my pronunciation of 鴨 "corrected" from /aap3/ to /ngaap3/.

GSJ said...

Michael (Regarding point 2): Wow, does that include 啱啱?

Michael said...

to GSJ - yes this does include 啱啱...although I can't imagine anyone saying 'aam1 aam1'...Maybe Richard Ho would ;)

Anonymous said...

I would say the pronunciations are lazy not because it is easier to speak them, but that children would get on with learning fewer consonants and/or vowels. With the 'n' and 'l' distinctions, kids have to learn two different consonants. By merging them into the single 'l' sound, of course, they need only learn one consonant.

The same with the rest. The result is spoken Cantonese with a less rich array of vowel/consonant repertory. That to me is pretty lazy.

GSJ said...

First of all, children still learn the /n/ consonant, which still appears at the end of syllables (coda position) as /n/ and not as /l/.

Second, did you read anything in this thread that was posted by people who actually know what they're talking about? The richness of a language's consonant inventory has to do with speakers' laziness--are you really that comfortable with just making stuff up? You should be embarrassed!